Introduction

Any geographical location on Earth can be accurately represented by co-ordinates, which contain two components, longitude and latitude, as seen in Fig. 1. Latitude and longitude are measured in degrees, minutes, and seconds. The equator acts as the reference line for latitudes, while the prime meridian is the reference line for longitudes1. The sharing of geographical co-ordinates is a primary element of military communication2.

The motivation for this research stems from the urgent demand for highly secure, covert, and real-time communication systems in modern warfare, where precise geographical co-ordinates play a pivotal role in mission planning, surveillance, and autonomous operations. As battlefield environments become increasingly digitised, adversaries employ advanced interception techniques, including signal monitoring and steganalysis, to compromise military networks. While previous works have explored cryptographic models and data hiding separately, recent surveys such as those by Zhang et al.3 and Banerjee et al.4 highlight that integrated systems combining encryption and steganography are still underexplored, especially for spatial data. Moreover, the lack of dedicated frameworks for co-ordinate-level information sharing in a military context leaves a substantial gap in the field.

In our initial investigation of existing tools and methodologies, we observed that most encryption models either incurred high computational costs or lacked resistance to brute-force attacks under constrained environments. Similarly, common steganographic models suffered from limited payload capacity or poor imperceptibility when tested against modern detection tools such as StegExpose and deep learning-based analysers4,5. In the context of tactical military communication, where latency, bandwidth optimisation, and untraceability are critical, none of the current solutions met all practical requirements. Recent work by Hossain et al.6 and Liu et al.5 emphasises the growing concern around secure transmission of geospatial intelligence, yet no standardised approach currently addresses end-to-end secure co-ordinate sharing with built-in key hiding mechanisms.

Therefore, this research is driven by the necessity to develop a lightweight, practical, and imperceptible solution that unifies robust encryption (AES) and advanced steganographic techniques (hash-based LSB and colour plane splicing), optimised for TCP/IP military communication channels. By embedding both encrypted data and the steganographic key within a sequence of images, including a specially curated key image, the proposed system eliminates vulnerabilities associated with traditional key exchanges and ensures security against visual inspection and computational attacks. In doing so, we aim to contribute a novel and actionable solution that aligns with the current needs of cyber-physical military infrastructure.

Despite the immense strategic value of geographical co-ordinate data, a detailed scrutiny of existing literature reveals a surprising gap: no publicly available system has been specifically designed for the secure, untraceable transmission of such sensitive information. Although various encryption and steganographic methods have been studied in isolation, their application to co-ordinate data remains unexplored. Through extensive experimentation, we observed that simple yet powerful combinations of information hiding techniques can meet the stringent security and practical implementation demands of military operations2,7. Building on this insight, we propose a co-ordinate transmission system that ensures confidentiality and integrity and prioritises untraceability and imperceptibility in hostile environments. Hence, the main goal of this paper is to fill this critical gap by constructing a novel, end-to-end, untraceable system for the secure transmission of co-ordinates, and to demonstrate its effectiveness and real-world applicability through rigorous testing and analysis.

This research is paramount for critical military operations, offering a wide-ranging impact in various strategic areas. By designing a robust system that ensures security and imperceptibility in transmitting sensitive co-ordinates, we aim to aid in multiple military activities, including offensive manoeuvres by ensuring accurate target information and surveillance operations by tracking enemy movements to gather intelligence. Similarly, military leaders can maintain situational awareness and manage troops across various units under their command through the precise and secure sharing of bulk location data. Figure 2 illustrates various military operation scenarios that rely on co-ordinate data, demonstrating how our transmission system can ensure the success of these missions.

Over time, the secret transmission of data has been a field of extensive research owing to its applicability, particularly in defence research2. Out of the many approaches to secret transmission, steganography and cryptography are the most used. As steganography ensures critical data is embedded into digital messages that do not seem necessary and, therefore, do not come under the scrutiny of potential hackers, it is selected to be used in our system.

Steganography is generally of two types8, text-based or image-based, depending on the initial data in which the secret information is then encoded. It is usually accompanied by encryption to ensure that only an authorised person can access the encoded secret data9. Symmetric key encryption, in which access to the data requires using the same key irrespective of the action performed (encryption or decryption), is often favoured over asymmetric key algorithms that use two keys: one public and the other private, due to simplicity and speed. A more detailed account of further terminology and key concepts used in our system is given in “Terminology and key concepts”.

Fig. 1
figure 1

Illustration of geographical terminology using the co-ordinates of Mumbai as an example.

Full size image
Fig. 2
figure 2

Military application scenarios for our proposed system.

Full size image

In this paper, we propose a system for secure transmission of co-ordinates by combining two standard information security techniques: encryption and steganography. The proposed system ensures that service personnel can safely share sensitive location data with minimal threat from hackers.

The core novelty of the proposed system lies in its simple yet efficient implementation with a streamlined, self-contained design that delivers complete security without reliance on external components or auxiliary mechanisms. Moreover, we also provide secure and innovative handling of steganographic keys, which are used as hash keys for pixel selection. Traditional key exchange methods, such as physical transfer or unsecured offline storage, introduce significant vulnerabilities to eavesdropping and interception. In contrast, our approach embeds the stego key directly within one of the transmitted images, enabling only the intended recipient to identify and extract it deterministically. This integrated mechanism eliminates the need for separate key exchange and enables safe and rapid transmission of sensitive co-ordinate data in high-risk environments.

The main contribution of this paper can be summarised as follows:

  • We identify and analyse the shortcomings of existing approaches and systems for secure transmission of co-ordinates in a military context.

  • We propose and develop a simple end-to-end architecture for the secure transmission of co-ordinates, combining information security techniques, including AES (Advanced Encryption Standard) encryption and multi-image steganography using hash-based LSB substitution and colour plane splicing.

  • We showcase the system’s efficiency, stability, and suitability for real-world applications through experimentation and rigorous testing on multiple image sizes and co-ordinates, with promising results.

A review of related published work, along with their potential drawbacks for secure transmission of co-ordinates is provided in the next section. In the subsequent sections, we present the algorithms used, encompassing the encryption and steganography processes involved in the research, and discuss the results obtained. Finally, we present the strengths and limitations of our system along with ideas for future improvements.

Literature survey and related works

Review of information security techniques used in the system

A high level of data confidentiality can be achieved through encryption. Multiple encryption techniques with their advantages have been surveyed in10. As stated in11, the Advanced Encryption Standard (AES) offers the highest level of security due to its large key size, making it a suitable choice for the encryption process in our system.

Implementing steganography techniques, or steganographic schemes, provides a high level of security to resist brute force attacks. Various image steganography techniques have been proposed and developed depending on transmission requirements. Steganography techniques built over the past three decades have been analysed in depth in12, in which the authors also provide a roadmap for using steganography tools. The LSB (i.e., least significant bits) replacement method is one of the simplest methods used to modify the LSB of cover pixels to embed secret data. Several variations of LSB encoding schemes have been suggested and implemented in recent times13,14,15. The LSB-based methods have thus been the most analysed and discussed steganography techniques. Detection techniques for the corresponding methods have also been improved and implemented for grayscale and colour images16.

Apart from LSB-based encoding, significant research has been done on using stego keys for image steganography. Recently, a technique that encodes text in a picture using up to the four least significant bits (LSB) based on a hash function has been proposed and proven effective17. Jain et al.18 and Karim et al.19 use stego keys, keys meant explicitly for use in the process of steganography, an idea that has been expanded in this paper. Using multiple stego-images has also been suggested to improve communication security, especially for the military20.

Along with steganography techniques for grayscale images, other novel steganography approaches in RGB colour images have also been suggested due to the prevalence of coloured media that is shared over networks21,22. Singh et al. suggest embedding secret data in RGB colour planes in the ratio 2:2:422, as the sensitivity of the human eye towards red and green components is more as compared to the blue component based on research conducted by Schubert23, and Juneja et al.24. Colour plane splicing has thus been recognised as an effective technique for increasing steganography security.

Thus, up-to-date cybersecurity techniques, including AES encryption, steganography using hash-based LSB substitution and bit encoding after colour plane splicing, have been utilised to develop our proposed system to ensure complete secrecy and maximum untraceability.

Review of related existing systems

Even though multiple secure communication technologies that facilitate the exchange of various types of data exist in military contexts, none specifically cater to the unique requirements and stringent security demands associated with the transmission of co-ordinates.

Table 1 summarises the main approach and drawbacks of the most relevant research in building a secure system for data exchange, focusing on work that can contribute to implementing techniques utilised in the system, namely encryption with multi-image steganography using hash-based LSB substitution and colour plane splicing. Drawbacks mentioned in the resources or identified through suitability analysis have also been included in the table to highlight the difficulty of using existing techniques and systems directly for our requirements. Our system aims to overcome these drawbacks to present a modern system designed specifically for the secure transmission of co-ordinates.

Table 1 Review of related existing systems and their weaknesses/drawbacks.
Full size table

Recent advancements and state-of-the-art comparison

More recently, there has been substantial improvement in secure techniques for transmitting data, especially for military and mission-critical purposes. Chen et al. (2024)30 proposed an encryption-based steganography system with deep auto-encoders for drone communications, but it did not prove to be perceptually unnoticed when tested using networks with limited resources. Likewise, Fang and Li (2025)31 presented a battlefield IoT network-optimised hybrid protocol for cryptography, which failed to adopt image-based covertness, leaving it open to visual eavesdropping attacks. On the steganalysis side, Verma et al. (2025)32 investigated adversarial learning for satellite image-based detection of LSB-based steganography, illustrating how classical single-image hiding schemes are being made ever-more vulnerable to attacks using artificially intelligent detection methods. These advancements highlight the need for multi-layered, image-based, and stealth-maximising schemes. By encrypting payloads across multiple channels for a given image and internal key exchange, our proposal harmonises with the mentioned constraints and proves resilient even with heavy statistical and artificially intelligent attacks.

Moreover, while existing methods mentioned above may perform well within specific domains, they often fall short of offering a holistic solution that combines payload scalability, imperceptibility, and resistance to detection. Additionally, reliance on simple key exchange mechanisms and single-image carriers makes them unsuitable for high-security applications such as transmitting military co-ordinates. In contrast, by using a hash-based pixel selection technique, our approach employs a multi-image framework that distributes encrypted spatial data across several carrier images. Crucially, the steganographic key is securely embedded within a designated key image, eliminating vulnerabilities associated with traditional key exchange. To better contrast the strengths and weaknesses of previous methods and to mark the novelty of our proposed scheme, Table 2 provides a state-of-the-art comparison across pertinent dimensions.

Table 2 State-of-the-art comparison of secure data transmission approaches.
Full size table

Proposed methodology

Terminology and key concepts

To ensure conceptual clarity, this section outlines the fundamental theories and concepts that form the backbone of the proposed system. These include key principles in steganography, cryptography, secure transmission, and their real-world applicability, especially in military-grade information exchange.

a. Steganography: Steganography is the science of hiding information within innocuous media such that its presence is concealed from unintended recipients. Unlike cryptography, which protects the content of the message, steganography hides the existence of the message itself. This dual use becomes especially powerful when used in tandem with encryption. In our system, steganography ensures encrypted co-ordinate data remains undetectable within image files.

b. Image steganography: A technique that conceals secret data within digital images by subtly altering pixel values. Its primary objective is imperceptibility, which ensures that the modified (stego) image remains visually and statistically indistinguishable from the original. This method is widely favoured for secure and covert communication because of the high redundancy and ample payload capacity of images.

c. Multi-image steganography: A variant of the image steganography technique that distributes the hidden message across multiple carrier images instead of a single one. This significantly enhances security by reducing the risk of complete data loss or detection. In the proposed system, the actual encrypted data, along with associated keys, is spread over a sequence of images, ensuring robustness and confidentiality even under partial data compromise.

d. Hash-based LSB steganography: The Least Significant Bit (LSB) steganography technique involves altering the least significant bits of image pixel values to store hidden information. We use a hash-based approach to dynamically select pixels for embedding based on a generated hash key to add randomness and security. This prevents sequential embedding patterns, improving resistance to steganalysis attacks. The key advantage lies in its simplicity, high PSNR (Peak Signal-to-Noise Ratio), and compatibility with existing image formats.

e. Colour plane splicing: This method builds on image steganography to embed secret data within the red, green, and blue (RGB) channels of an image. Human vision is more sensitive to changes in the red and green planes, while the blue plane can accommodate more significant alterations without perceptual detection. Our system leverages this property to embed different data bits in different colour planes, improving imperceptibility and increasing the embedding capacity.

f. Cover image: A cover image is the original digital image into which secret data is embedded. The choice of cover images in our system is crucial; they are selected for high texture variation and low semantic relevance to avoid drawing suspicion. Cover images are the medium for secure communication without revealing any trace of hidden content.

g. Stego image: After embedding the encrypted data, the modified cover image is referred to as the stego image. A good stego image should retain visual similarity to the original and withstand common image processing operations such as compression, scaling, and slight noise addition.

h. Hash key: The hash key used in our system is a three-part composite key that includes a modulus, an offset, and a colour plane identifier. It is deterministically generated using a secure algorithm and guides the pixel selection process during steganography. This key is securely embedded within a special key image to facilitate extraction and reconstruction at the receiver’s end.

i. Encryption: Encryption is a method of converting plain-text into cipher-text using a cryptographic algorithm and a key, ensuring that only authorised recipients can access the original data. It provides data confidentiality and is the first line of defence in information security.

j. Advanced encryption standard (AES): AES is a symmetric-key block cipher standardised and widely adopted due to its balance of efficiency and security. It operates on 128-bit blocks with key sizes of 128, 192, or 256 bits. AES performs multiple rounds of substitution, permutation, and mixing input data based on the key, making it highly resistant to brute-force and known plain-text attacks. In our system, AES is used to encrypt the co-ordinates before embedding, adding a critical layer of protection.

k. Symmetric vs. asymmetric encryption: Symmetric encryption uses the same secret key for encryption and decryption, making it fast and resource-efficient. Asymmetric encryption uses a public and private key pair and is more computationally intensive but offers better scalability for open networks. Since our system is designed for secure transmission over trusted channels (e.g., authenticated military nodes), symmetric encryption (AES) is preferred for its speed and simplicity.

l. Secure channel: Data is transmitted over a TCP/IP channel enhanced with SSL/TLS security protocols. SSL/TLS ensures message integrity and confidentiality by encrypting data packets and establishing a secure handshake before transmission. This is critical in military communication, where interception is a severe risk. Owing to the widespread use of such channels, our system assumes the presence of such a secure communication layer during the image transmission phase.

Fig. 3
figure 3

Sender side flowchart.

Full size image
Fig. 4
figure 4

Receiver side flowchart.

Full size image

Proposed system architecture

The encryption cum hash-based LSB steganography process we use is described in detail by the flowcharts in Figs. 3 and  4. The flowcharts illustrate the entire process of using the secure transmission technique to share co-ordinate data as part of military communication.

As illustrated in the sender-side architecture (Fig. 3), the co-ordinate data, extracted manually or automatically from a map, is first encrypted using the AES algorithm. The resulting cipher-text and the AES encryption key are then passed to the embedding module. This module accesses a pre-identified database of cover images and employs a hash-based pixel selection strategy to embed the encrypted co-ordinates securely, producing a sequence of stego images. Additionally, the AES encryption key and the hash key used for embedding are embedded into a separate stego image, referred to as the key image, which is appended to the stego image sequence. The stego images are then transmitted over a secure TCP/IP network.

On the receiver side (Fig. 4), the final image in the received sequence (the key image) is used to extract both the encryption and hash keys. Due to the connection-oriented nature of the network, maintaining the order of image transmission is crucial. Using the retrieved hash key, the receiver extracts the cipher-text from the stego images and subsequently decrypts it using the AES key to recover the original co-ordinate data. This completes a fully secure, self-contained, and imperceptible end-to-end transmission process.

While the combination of encryption and steganography may appear conventional, the originality of this research lies in integrating multiple enhanced security layers specifically tailored for the secure transmission of co-ordinate data in military environments. A common critique of Least Significant Bit (LSB) steganography is its susceptibility to statistical steganalysis; however, our approach deviates from traditional LSB embedding by incorporating a hash-based pixel selection mechanism. This method employs a 3-component hash key to randomly and deterministically select pixel positions across multiple cover images, thereby reducing spatial predictability and rendering the embedding process robust against detection. In addition, our system performs colour-plane splicing based on perceptual colour sensitivity, distributing the payload adaptively across red, green, and blue channels to maximise imperceptibility while preserving visual fidelity. Notably, using a dedicated ’key image’, which embeds both the AES encryption key and the hash key, eliminates the need for external key exchange, a recognised weakness in existing systems. Combining these innovations ensures a secure, scalable, and imperceptible transmission pipeline suitable for real-world defence communication. Our strategy aligns with emerging steganographic models such as the one proposed by Luo et al. (2024)33, who demonstrate that LSB-based systems, when equipped with adaptive hashing and multi-carrier redundancy, remain highly effective for secure information hiding. Thus, our work not only reinvents LSB within a secure communication framework but also addresses long-standing challenges in covert data sharing with measurable enhancements in robustness, efficiency, and stealth. We also acknowledge that while message compression is a known optimisation that we can use for reducing payload size, it is not included in our system as it introduces additional linear computational complexity at both ends, whereas our design prioritises efficiency, simplicity, and scalability to handle large-scale communication without added overhead.

To further strengthen the system’s response to such advanced threats, once a cropping or replay attack is detected (via pixel mismatch or invalid session identifiers), the receiver-side algorithm immediately aborts the current decoding session and triggers a tampering alert. This alert can optionally initiate a secure re-request of the image set from the sender through a separate authenticated channel, or log the incident for further manual or automated review, depending on the operational protocol in use. These recovery strategies can be integrated with session metadata or blockchain-backed audit trails in future deployments. Thus, our system not only prevents successful decoding of tampered content but also provides a procedural pathway for identifying and addressing such threats in real-time military-grade communication environments.

While the system could theoretically benefit from message compression to reduce embedding payload, we have deliberately excluded compression techniques in the current design to preserve computational simplicity and operational robustness. Compression introduces non-trivial processing overhead and potential signature patterns that could compromise the stealth of the transmission-both of which are undesirable in time-sensitive and adversarial military scenarios. Instead, scalability is achieved by leveraging the dynamic flexibility of our hash-based pixel allocation mechanism, allowing users to embed more bits per image when needed, without affecting visual imperceptibility. Furthermore, with the abundance of publicly available or synthetically generated cover images, the system can scale horizontally by increasing the number of stego images, maintaining a lightweight and practical deployment model. This trade-off ensures that the system remains efficient, secure, and adaptable for real-world secure communication applications.

To ensure strong untraceability, the proposed system is designed to regenerate both the AES encryption key and the hash-based steganographic key for each transmission session. While the hash key is dynamically generated using Algorithm 3, the AES key is also freshly generated for every new set of co-ordinate data. This session-level regeneration ensures that every stego image sequence is unique, preventing correlation attacks across multiple transmissions. As no fixed credential or persistent identifier is embedded or reused, even an attacker with full access to previous images or device memory cannot link messages across sessions. This design decision reinforces the system’s capability to maintain confidentiality, unlinkability, and resilience under strong adversarial models.

Process explanation

The complete process for transmitting co-ordinate data in military communication using the proposed system is detailed in the following sections, along with corresponding algorithmic steps. “AES encryption” outlines the AES-based encryption procedure for securing the co-ordinate data. “Sender side embedding algorithm” presents the sender-side embedding algorithm, which incorporates the hash key generated as described in “Sender side hash key generator algorithm”. Finally, “Receiver side algorithm” details the receiver-side extraction and decryption process, completing the end-to-end secure communication framework.

AES encryption

Advanced Encryption Standard (AES) is a symmetric key block cipher method for encrypting data. Unlike asymmetric-key block ciphers, AES uses the same key for encryption and decryption. AES operates on fixed-size blocks of data, with a block size of 128 bits. The key length for AES can vary and is commonly chosen as 128, 192, or 256 bits. AES replaces the older Data Encryption Standard (DES), improving security and efficiency.

Fig. 5
figure 5

Structure of the AES algorithm9.

Full size image

In our proposed system, AES with a 128-bit key is utilised to encrypt the co-ordinates that need to be securely transmitted. The algorithm takes a fixed number of co-ordinates as input and returns an encrypted form of the same using an encryption key. AES does not directly encrypt a decimal point. To handle this, we propose the use of a ’position bit’ to indicate the position of the decimal point in the original latitude or longitude. The general structure of AES can be represented as shown in Fig. 5.

The structure of AES involves multiple rounds of substitution, permutation, and mixing operations. It operates on 128-bit data blocks, but the size of individual co-ordinate data, such as latitude or longitude, can be represented using fewer bits depending on the accuracy required. Hence, 3 of the leading zeroes at the start of the latitude or longitude chunk represent the decimal point by setting a ‘position bit’ that represents the number of digits after which a decimal point must be given. This means that if there is a ‘1’ in the 2nd position from the left in the input latitude or longitude, then a decimal point will be given after two digits in the extracted and decrypted latitude or longitude after removing the leading zeroes and position bit.

Additionally, to represent the sign of the co-ordinate, the fourth bit from the left in the plain-text block is used. A ’0’ in this position indicates a positive value, while a ’1’ denotes a negative value. This convention aligns with the valid ranges for latitude (\(\text {-}90^{\circ }\) to \(+90^{\circ }\)) and longitude (\(\text {-}180^{\circ }\) to \(+180^{\circ }\)).

For example, the co-ordinates of Mumbai are \(latitude: +19.075983\) and \(longitude: +72.877655\). These are represented as 0100000019075983 and 0100000072877655, which are provided as inputs to the AES algorithm.

Algorithm 1
figure a

AES-based encryption of co-ordinates.

Full size image

This algorithm securely encrypts a list of geographical co-ordinates using the Advanced Encryption Standard (AES). Each latitude and longitude value is separately encrypted using a symmetric AES key. The encryption process follows standard AES steps across multiple rounds. The function AES_Encrypt encapsulates this process for individual values. The main function iterates through each co-ordinate pair, encrypts both components, and stores them in a new list. This ensures confidentiality and prevents unauthorised access to sensitive location data.

Sender side embedding algorithm

Embedding refers to the process of using image steganography to insert secret data into pixels of innocuous-looking images to transmit it without suspicion or detection by attackers. The sender-side embedding algorithm uses a modified hashing-based LSB method to embed binary co-ordinate data. The algorithm generates a set of stego images and a key image that holds the fundamental hashing key and encryption key.

The process first converts the encrypted co-ordinates to binary format, then calculates the hash key, which dictates the number of cover images required and then splices them into RGB colour planes. The hash key is then used to select the pixels for steganography and the colour plane in which data is embedded. After completion of the algorithm process, all the stego images and the key image are sent over a secure TCP connection to maintain the sequence of all stego images first and then the key image.

Algorithm 2
figure b

Sender-side embedding.

Full size image

This algorithm outlines the steganographic embedding of encrypted co-ordinate data into multiple cover images using a hash-based LSB technique. The secret message is converted into binary and padded to match the number of embedding pixels. A custom hash key is generated, which guides the pixel and colour channel selection for embedding each bit. Red, green, or blue channels are modified depending on the hash key’s value to minimise perceptual change. Additional adjustments are applied to the blue channel to preserve image quality. The 22-bit hash key and 128-bit AES key are embedded into a designated key image for secure retrieval. Finally, the complete stego images, along with the key image, are transmitted over a secure TCP channel, ensuring both confidentiality and imperceptibility.

Sender side hash key generator algorithm

The hash generator algorithm, the fundamental algorithm used for selecting pixels from cover images for steganography, produces a hash key and the number of pixels to be used for embedding per cover image. This is helpful in determining the number of cover images required. It takes into account the fact that the sender and receiver have mutually agreed on the hashing function to be used.

The hash key and count of pixels generated by the hash generator have the following properties:

  1. (1)

    The hash key consists of 3 values, the first two of which act as parameters for the sender side embedding algorithm, and the third value will be used to decide the colour plane (RGB) in which data is to be embedded in the selected pixel.

  2. (2)

    The hash key values guarantee that the count of pixels to be used for embedding is within a specified range and can successfully embed the required number of co-ordinates in the number of cover images selected, with minimal (much less than 5% of message size) addition of redundant zeroes at the end of the co-ordinate bits.

  3. (3)

    The first two hash key values can be specified to lie within a particular range. It has been found by extensive mathematical experimentation that at least one pair of values exists between a range of 280 to 300, which can successfully embed anywhere between 5 to 40 co-ordinates at a time.

A range of 280 to 300 is thus given as a proven default range to the hash generator; however, mathematical research detailing the ideal hash function with an ideal range of values can lead to further interesting results.

Algorithm 3
figure c

Hash key generator.

Full size image

This algorithm generates a secure and randomised hash key for pixel selection in the steganographic process. It takes the number of co-ordinates, cover image dimensions, and a predefined range to identify suitable modulus-offset pairs that determine which pixels will carry the hidden data. The algorithm iterates through the given range to compute pixel positions satisfying the mathematical embedding condition and filters them based on a valid pixel count range and message divisibility. The selected tuple provides the hash modulus and offset, while a random colour channel selector adds variability. This dynamic selection enhances both the robustness and imperceptibility of the embedding process by ensuring unpredictability in pixel and channel mapping.

Receiver side algorithm

The receiver side algorithm follows a mirror process of the algorithm followed at the sender side. The receiver accepts a series of stego images that are sent over a TCP connection by the sender, maintaining their sequence. The receiver side algorithm first identifies the last image as the key image to extract the hash key and encryption key used for the decoding process ahead. It then follows the same process on the sender side by using the mutually agreed-upon hash function and received hash key to identify the pixels in which data has been embedded and constructs a message stream of encoded co-ordinates. It then divides the message stream into 256-bit chunks, converts them to hexadecimal format and finally applies the received encryption key to generate a set of decoded co-ordinates, thus completing the highly secure communication process. Interestingly, in this process, the knowledge that the hash key and encryption key are retrievable only from the last image in the sequence is exclusive to the receiver. This significantly reduces the likelihood of a potential hacker obtaining access to these critical keys. Thus, despite using hashing, the symmetric key transmission remains secure, as the method ensures that only the intended recipient can access the keys needed for decryption.

Algorithm 4
figure d

Receiver side algorithm.

Full size image

This algorithm is responsible for securely extracting and decoding co-ordinate data from a received set of stego-images and a key image. It begins by retrieving the embedded hash key and AES encryption key from specific pixels in the key image. Using these values, it traverses each stego-image to extract the least significant or specific bits of selected pixels based on a hashing condition. The collected bitstream is split into 256-bit segments, which are decrypted using AES to obtain the original binary representation of latitude and longitude. A positional decoding strategy is used to restore decimal values by interpreting embedded prefix bits. The final output is a list of geospatial co-ordinates recovered accurately from the encrypted and concealed transmission. As an additional security check, a mismatch between expected and extracted pixel count based on the shared hash key is also identified and serves as a strong indicator of tampering or image manipulation like cropping, which can be flagged.

Computational costs of the algorithms

This section analyses the computational time complexity of the algorithms outlined in “AES encryption” to “Receiver side algorithm”, using standard asymptotic notations as defined in34. The system incurs no variable memory overhead, as the only storage requirement-image storage-remains constant. Let C1 to C4 represent the computational costs of the respective algorithms, and Ctotal denote the total cost of end-to-end plaintext message transmission from sender to receiver.

  • Computational cost of Algorithm 1: encryption of co-ordinates using AES In general, the cost of encryption can be approximated as the product of the number of co-ordinates (M) and the computational cost of encrypting each co-ordinate using AES, which is constant since AES operates on a fixed block size and takes approximately the same time independently of input.

    $$\begin{aligned} C_1= & O(M) \end{aligned}$$
    (1)

  • Computational costs of Algorithm 2: sender side embedding algorithm and Algorithm 3: hash key generator algorithm The computational cost of the hash key generation algorithm depends on static input parameters: min_val, max_val, and the dimensions of the selected image (in pixels). This results in a cost that grows linearly with the size of the image:

    $$\begin{aligned} C_3= & O((\texttt {max\_val} - \texttt {min\_val}) \times (\texttt {width} \times \texttt {height})) \end{aligned}$$
    (2)
    $$\begin{aligned}= & O(k) \end{aligned}$$
    (3)
    $$\begin{aligned}= & k \cdot O(1) \end{aligned}$$
    (4)

    where k denotes the total number of pixels in the image. The cost of the sender-side embedding algorithm is determined by the number of images to be embedded (N) and the embedding cost per image (proportional to its fixed resolution):

    $$\begin{aligned} C_2= & O(\texttt {width} \times \texttt {height} \times N) + C_3 \end{aligned}$$
    (5)
    $$\begin{aligned}= & O(k \cdot N) \end{aligned}$$
    (6)
    $$\begin{aligned}= & k \cdot O(N) \end{aligned}$$
    (7)

    Accordingly, the total computational cost on the sender side, including encryption (denoted by \(C_1 = O(M)\) for message size M), embedding, and network transmission, is given by:

    $$\begin{aligned} \texttt {Sender\_Side\_Cost}= & C_1 + C_2 + \texttt {Network\_Cost} \end{aligned}$$
    (8)
    $$\begin{aligned}= & O(M) + O(k \cdot N) + \texttt {Network\_Cost} \end{aligned}$$
    (9)
    $$\begin{aligned}= & O(M) + k \cdot O(N) + \texttt {Network\_Cost} \end{aligned}$$
    (10)

    Hence, the sender-size cost depends linearly only on the parameters of the number of co-ordinates (M) and the number of images to be embedded (N), with additional network costs, if any.

  • Computational cost of Algorithm 4: receiver side decoding and extraction algorithm The receiver performs a mirrored process to extract and decode the co-ordinate data. As the network transmission cost is accounted for on the sender side, the receiver-side cost mirrors the computational components of the sender-side cost, excluding the network overhead:

    $$\begin{aligned} \texttt {Receiver\_Side\_Cost}= & \texttt {Sender\_Side\_Cost} - \texttt {Network\_Cost} \end{aligned}$$
    (11)
    $$\begin{aligned}= & O(M) + O(\texttt {width} \times \texttt {height} \times N) \end{aligned}$$
    (12)
    $$\begin{aligned}= & O(M) + k \cdot O(N) \end{aligned}$$
    (13)

  • Total message transmission cost The total computational cost for securely transmitting a list of co-ordinates from the sender to the receiver in plaintext form is therefore:

    $$\begin{aligned} C_{\text {total}}= & \texttt {Sender\_Side\_Cost} + \texttt {Receiver\_Side\_Cost} \end{aligned}$$
    (14)
    $$\begin{aligned}= & 2 \cdot (O(M) + k \cdot O(N)) + \texttt {Network\_Cost} \end{aligned}$$
    (15)

    When using a secure TCP/IP channel, network cost is further optimised. For a fixed image size (constant k), the total computational cost grows linearly with both the number of co-ordinates (M) and the number of stego images required for embedding (N). This demonstrates the system’s practical algorithmic efficiency under real-world communication scenarios.

System requirements

After a careful study of official military communication requirements2, we detail the key security and implementation requirements for a system designed for our purpose in military operations in this section. A practically deployed system must ensure top-notch security while balancing practicality at all times.

Implementation requirements

Robustness: Capable of easily handling rapid and large-scale communication

Simplicity: Easy to operate in challenging field environments with minimal human intervention

Efficiency: Ensure that the system can function in resource-constrained environments

Security requirements

Following in-depth scrutiny of publicly-released military requirements2,7, we identify the following security requirements for a secure co-ordinate transmission system. We use these requirements to build our threat model and detail the countermeasures for protection against the identified attack vectors.

  • Confidentiality: Prevent unauthorised access at any point in communication

  • Integrity: Ensure the data remains unaltered at all transmission stages

  • Untraceability: Conceal the very existence of data transmission

  • Imperceptibility: Ensure minimal change in innocuous data to minimise suspicion

Threat model

To ensure the resilience of the proposed system against potential security threats, we present a comprehensive threat and attack model. The threat model outlines various adversaries with different capabilities, resources, and motivations. By correlating each identified threat with the key security requirements, we design appropriate countermeasures to ensure the resilience of the system.

Attackers and attack vectors

The proposed system for military communication can have multiple adversaries with diverse levels of knowledge, resources, and motivations. Their goal is to breach confidentiality and imperceptibility. We acknowledge the presence of the following potential attack vectors from these attackers:

  1. (1)

    Passive attackers: Attackers with lower knowledge and resources who attempt to eavesdrop and intercept sensitive co-ordinate data without alteration. Their attack vectors can include:

    1. (a)

      Network eavesdropping: Attempts to identify data packets sent over the TCP/IP channel, compromising confidentiality and untraceability

    2. (b)

      Data interception: Attempts to intercept and extract co-ordinate data during transmission, affecting integrity and confidentiality

  2. (2)

    Active attackers: Attackers with high knowledge, motive and resources to modify the transmitted data, compromising the integrity and confidentiality of the communication. Their attack vectors can include:

    1. (a)

      Data manipulation: Identifies data manipulations made, harming imperceptibility, and then attempts to actively re-manipulate the data to compromise its integrity and authenticity

    2. (b)

      Replay transmissions: Attempts to capture and retransmit previously intercepted data packets to disrupt the communication flow, targeting integrity and also untraceability

    3. (c)

      Cropping attacks: Attackers may attempt to crop steganographic images or replay previously captured packets

Types of attacks and countermeasures

The various types of attacks the system may encounter, as well as the security measures incorporated into the system to mitigate these attacks, are presented in Fig. 6, showing how we satisfy every security requirement detailed in the previous section. Even though security is the top priority, our system also balances it with implementation requirements to ensure practicality at all times. In case of advanced attacks, we have explained in Algorithm 4 how data manipulation or cropping will result in the loss of embedded pixels, which can be mitigated in our system using a hash-key-based pixel validation mechanism. Any deviation in the expected number of embedding pixels, computed via a shared deterministic hash function, triggers a tampering alert and prevents further decryption. This ensures that modified or replayed content cannot be accepted as valid.

As stated before, our system does not have a predecessor to exactly compare to, hence we use detailed mathematical results to show the effectiveness of the system in information hiding through various experiments, described in Section 4. By integrating all possible countermeasures into the system design, we mitigate the identified attacks effectively, ensuring the secure transmission of co-ordinate data in extreme operational environments.

To further strengthen the system’s resilience against advanced attacks, we have incorporated targeted countermeasures for data manipulation attacks, replay attacks, and cropping attacks. Cropping and data manipulation are detected through our hash-based pixel validation mechanism, which ensures that any deviation in the expected number of embedding pixels triggers a tampering alert and halts further decryption. Replay attacks are mitigated through per-session regeneration of both the AES encryption key and the hash key, ensuring that each transmission produces a cryptographically unique stego image set. As a result, intercepted images from previous sessions cannot be reused or correlated. Furthermore, since no static identifiers or reusable credentials are embedded, the system maintains strong untraceability even under adversarial models with full access to device memory or transmission history. These defenses are built into the core architecture and have been rigorously validated, ensuring that the proposed protocol is robust against both conventional and advanced threat vectors in real-world deployment scenarios.

Fig. 6
figure 6

Types of attacks and the corresponding countermeasures implemented in the system.

Full size image

Experimentation and result discussion

Since we identify no existing communication system dedicated to the communication of co-ordinates for the military, we instead design several scrutinising experiments involving both mathematical and programmatic attack techniques to prove that our system satisfies all the security and implementation requirements, described ahead. To simulate practical scenarios, a total of 100 tests with varying data were carried out, and the results obtained were then tallied against benchmark results obtained from standard security algorithms. Analysis of all experiments conducted can be found in “Attackers and attack vectors”, while a particular sample test case has been described and analysed in “Implementation requirements”, “Security requirements” and “Threat model”.

Fig. 7
figure 7

Illustrative example of input data and the sender-side workflow in the proposed system.

Full size image

To demonstrate the process used in this paper, the sample test case along with the data utilised and processing steps at the sender side is illustrated in Fig. 7. In the figure, step 1 shows the sample input data, including the number of co-ordinates to be embedded and the corresponding set of co-ordinates. In step 2, each co-ordinate is then represented as a 16-character plain text by setting the ‘position bit’ and ’sign bit’ as explained in “AES encryption”. As an example, the representation of the co-ordinates for Pune has been given, where sign indicators (4th position from left) for latitude and longitude are 0 indicating a positive sign for both, and the latitude and longitude both have a value of 1 set in the 2nd position from the right, indicating that a decimal point should be given two digits after removing leading zeroes, position and sign indicators.

In step 3 of the figure, the co-ordinates are then encrypted using the AES algorithm with a hexadecimal key as shown. Each pair of encrypted co-ordinates is then converted to a 256-bit binary format and concatenated to form the input of length 1536 bits for the steganography process as given in step 4.

Finally, in step 5 shown in the figure, the hash key generator is used to get a suitable hash key and count of pixels to be embedded per stego image, which dictates that the number of stego images to be used was 2. The open-source images ‘lena.jpg’ and ‘mandril.jpg’ are used as cover images to embed the co-ordinates, and the ‘airplaneF16.jpg’ image is used as a key image, all of size 512 x 512 as shown in Fig. 8.

After following this process, the performance metrics for the system requirements outlined in “Practical simulation” were calculated and are presented in the following sections.

Fig. 8
figure 8

Top row represents cover images while the bottom row shows images embedded with the secret co-ordinates and keys.

Full size image
Table 3 Histogram analysis.
Full size table
Fig. 9
figure 9

Histograms of original images in the top row compared with histograms of images with embedded data in the bottom row.

Full size image

Mathematical security analysis

In this section, we use mathematical and statistical techniques to show how our system satisfies the security requirements of confidentiality, integrity, imperceptibility and untraceability.

To validate the security of the proposed system, a formal analysis was conducted using both mathematical models and programmatic simulations, addressing key requirements such as confidentiality, integrity, imperceptibility, and robustness. Mathematical models, including histogram comparison, Mean Square Error (MSE), and Peak Signal-to-Noise Ratio (PSNR), were used to evaluate the similarity between cover and stego images, confirming minimal distortion and ensuring imperceptibility and data integrity. Programmatic models involved the use of advanced steganalysis tools like StegDetect and StegExpose, which reported no detection of embedded data, thereby validating the system’s untraceability. Additionally, network-level tests assessed the system’s behaviour under constrained transmission scenarios, while practical simulations were conducted through 100 test cases with varied image sizes to evaluate consistency, payload scalability, and real-world applicability. These results collectively confirm that the system meets the stringent security standards required for covert military communication.

Histogram analysis

The chief metric used for evaluation of similarity between images, the plotting of histograms, as shown in Fig. 9, shows minimal difference in the original images and the stego images, highlighting confidentiality. Standard histogram comparison metrics suggested by the OpenCV documentation35 have been computed and tabulated in Table 3. According to35, for the correlation method, the higher the metric, the more accurate the match, while the reverse is true for the Chi-square and Bhattacharya methods. A high score obtained in each of these metrics thus proves effective in displaying the efficiency and imperceptibility of the proposed process used for steganography.

MSE (mean square error) and PSNR (peak signal-to-noise ratio)

MSE and PSNR are the primary metrics used to evaluate steganographic performance and image fidelity16. MSE quantifies the average squared pixel-wise difference between the cover image and the stego image. PSNR, as detailed in36,37,38, serves as a standard measure of image quality by comparing the peak signal power to the distortion introduced. The formulas used for computing MSE and PSNR are given below.

$$\begin{aligned} \text {MSE} = \frac{1}{mn} \sum _{i=1}^{m} \sum _{j=1}^{n} (x_{ij} - y_{ij})^2 \end{aligned}$$
(16)
$$\begin{aligned} \text {PSNR} = 10 \cdot \log _{10} \left( \frac{R^2}{\text {MSE}} \right) \end{aligned}$$
(17)

Here, m and n denote the width and height of the image, x and y represent the pixel values of the cover and stego images, respectively, and R is the maximum possible pixel intensity (255 for 8-bit images, as only one RGB channel is altered).

To assess the effectiveness of the proposed system, PSNR values were computed for all stego images. A higher PSNR indicates better imperceptibility and image quality. The calculated MSE and PSNR values are summarised in Table 4.

Table 4 Analysis of stego images and key image.
Full size table

The PSNR value of the hash-based LSB technique can be plotted against existing techniques for ‘lena.jpg’ as given in Table 5. It has been proven in39,40 that humans cannot perceive the difference in the images if PSNR is greater than 36 dB. As all images used for experimentation effectively satisfy this criterion, it can be concluded that the efficiency of the proposed hash-based LSB technique is significant for this particular use case.

Table 5 Comparison of the hash-based LSB method with other LSB-based methods.
Full size table

Programmatic security analysis

Here, we simulate the system and use programmatic attacks to highlight the simple yet robust security measures in our system.

Steganographic attacks

In order to emphasise the imperceptibility and untraceability of the system, steganalysis, the process that tries to defeat steganography by detecting hidden data, was conducted using state-of-the-art open-source software StegDetect45 and StegExpose46, techniques for which were last updated in 2023. StegExpose uses an algorithm derived from an intelligent and thoroughly tested combination of preexisting pixel-based steganalysis methods, which has been thoroughly tested on over 15000 images. StegDetect is a Python-based tool that is focused on the detection of steganography, specifically using the LSB technique.

Steganalysis using both tools did not produce suspicious pixels in any image generated using this process, as tabulated in Table 6, which provides substantial support to the security of the proposed steganographic process.

Table 6 Number of suspicious pixels detected by steganalysis.
Full size table

Network attacks

To address the threat of passive network attacks such as eavesdropping or traffic analysis, we can directly attribute the light network footprint of our system, which ensures inconspicuous communication. If we consider approximately 900 usable pixels per image and each co-ordinate requiring 256 bits, transmitting as many as 100 co-ordinates demands fewer than 30 images (around 300kb using compressed 10kb images), a volume well within the range of normal digital communication today, such as a small batch of social media uploads.

Practical simulation

To further demonstrate the efficiency and imperceptibility of our system and algorithms, we simulated a total of 100 experiments by practically running the system on two personal computers. A set of 300 random images with sizes as 256px x 256px, 512px x 512px, 1080px x 1080px or 1280px x 720px, obtained from the open-source image website Unsplash47, were divided into 100 sets of 3 images (2 stego images and 1 key image per set, all of same size) for testing. A randomly chosen set of 5 co-ordinates encrypted with an encryption key of ’2B7E151628AED2A6ABF7158809CF4F3C’ and embedded with an appropriate hash key generated using Algorithm-3 was then used for testing purposes on the selected images.

The variation and distribution of MSE and PSNR values for each stego image and the key image have been presented using line graphs and histograms. The entire distribution of results obtained from all 100 experiments has been collectively shown in Fig. 10 and Fig. 11.

The statistical summary of MSE and PSNR values for the set of experiments grouped by the size of images has been tabulated in Table 7.

Fig. 10
figure 10

Line plots showing the variation of MSE values (a–c) and PSNR values (d–f) for each image in the 100 experiments carried out to test the system as described in “Practical simulation”.

Full size image
Fig. 11
figure 11

Histogram plots for MSE values (a –c) and PSNR values (d–f) for the 100 experiments carried out to test the system as described in “Practical simulation”.

Full size image

The findings and inferences from the results obtained after the experimental testing can be presented as follows:

  • MSE values show a marked decrease, and PSNR values show a high increase as the size of images increases due to increased payload capacity. Thus, using larger size of images can ensure maximum imperceptibility in transmission using the system

  • For images of a particular size, the range of PSNR and MSE values (difference between maximum and minimum values) is minimum, showing stability and consistency in the performance of the entire system (groupings according to image size are also observed from the histograms in Fig. 11)

  • Using maximum image size for the key image can guarantee the imperceptibility of the keys embedded in the image

  • With increasing image storage sizes, transmission costs also increase proportionally. Our system ensures that, regardless of the size of images used, the criterion of PSNR value above 36 dB is satisfied with minimal overall MSE (even for low-quality images) as seen by the line plots in Fig. 10. This clearly showcases the real-world applicability of the system even for low bandwidth communication channels commonly seen in critical military conditions

Table 7 Maximum, minimum and average MSE and PSNR values for the 100 tests grouped by image size.
Full size table

Security against physical device compromise

An important real-world threat model is the possibility of a physical device compromise by a strong adversary who may gain access to internal files. In our system, such a scenario is carefully mitigated. The encryption and hash-based steganographic keys are not stored on disk or retained in any identifiable format; instead, they are dynamically embedded within a final key image. This image is indistinguishable from the rest of the stego images and not marked or ordered in any predictable way. Additionally, the key extraction process is governed by a hash-based pixel selection strategy, making it statistically infeasible to reverse engineer the keys even with access to all the transmitted images. As the key image location, its embedded bit lengths, and the hash function parameters are all unknown, brute-force recovery becomes computationally impractical.

Our approach aligns with modern research, which emphasises key obfuscation, randomised embedding, and passive attack resistance in steganographic frameworks48,49. Furthermore, the unpredictable and probabilistic nature of our hash-key generation logic introduces entropy comparable to modern zero-trust systems50, significantly increasing the resilience against device-level compromise. This design consideration reinforces the resilience of our system even in the event of full physical access by an attacker.

Conclusion and future scope

This research proposes a secure, end-to-end communication system for the imperceptible transmission of location-sensitive co-ordinate data, specifically designed for military applications. The proposed architecture uniquely integrates AES encryption with a hash-based LSB multi-image steganography technique, further enhanced through colour plane splicing and integrated key embedding. Unlike conventional approaches, this system avoids external key exchange by embedding both the encryption and hash keys within a designated key image, which is indistinguishable from other stego images.

Comprehensive experimental validation was performed through 100 simulations using images of various resolutions (256\(\times\)256 to 1280\(\times\)720), demonstrating high imperceptibility (PSNR values consistently above 80 dB), minimal distortion (MSE values < 0.001), and robustness against steganographic detection using tools like StegDetect and StegExpose, both of which detected zero suspicious pixels. These results validate the system’s effectiveness in maintaining confidentiality, untraceability, and operational resilience under real-world constraints.

In addition to satisfying standard information security metrics, this system also addresses critical threats such as physical device compromise by embedding keys in non-obvious, statistically unpredictable locations. These contributions, supported by algorithmic innovations and rigorous testing, underscore the novelty and significance of the work. The following subsections summarise the system’s key strengths, limitations, and potential future enhancements.

System strengths

  • Dual security: The system combines two of the most widely used security techniques, viz., encryption, and image steganography. It combines their benefit to create a military-grade end-to-end communication system for geographical co-ordinates. The overall security is enhanced since, even if an attacker identifies the steganographic image, they still need to decrypt the hidden data, which involves identifying the 128-bit AES encryption key. This redundancy provides multiple layers of protection so that even if one layer is compromised, the other still protects the data.

  • Integrated key management: The unique approach of embedding both the hash key and encryption key within the final stego image is a strategic adaptation that eliminates the need for third-party involvement or external communication channels in exchanging key data. Without the need to transmit keys separately, the risk of key interception is minimised. Additionally, it simplifies operations by avoiding the complexities of managing separate key exchanges and securing extra communication channels. Overall, this method underlines the system’s independence and self-sufficient nature.

  • Maximum imperceptibility: The system can ensure very high imperceptibility in transmission when using large-sized images. Higher imperceptibility means that the modifications made to embed data within the image are not noticeable to the naked eye or through basic analysis. As outlined in “Sender side embedding algorithm”, the number of pixels per image used for steganography is calculated during the embedding process. With larger resolution images, the chosen pixels are sufficiently spaced apart, making the stego image and the cover image appear identical. Additionally, with rapid advancements in digital image generation and sharing techniques, the system will become even more valuable in modern applications.

  • Low computational complexity: The algorithmic complexity of the system scales linearly with the number of co-ordinates and images being embedded, without additional overhead. For example, in a large-scale military surveillance operation tracking the real-time positions of numerous assets across a broad area, the system’s linear complexity ensures efficient handling of extensive data volumes. This capability minimises delays and performance issues, which is crucial for the success of the operation. Thus, the system maintains high computational efficiency in critical situations.

  • Practical applicability: Extensive testing of the system using images ranging from 256x256 to 1280x720 pixels, along with various co-ordinate sets, has yielded promising results. This rigorous evaluation proves our system’s reliability, stability, and practical suitability for real-world applications.

System limitations

  • Fixed application scope: The current embedding algorithms are specifically tailored for sharing location co-ordinates in military contexts, limiting their applicability. To broaden the system’s use, adaptations are needed to handle various sensitive data formats, such as medical records and financial data. For instance, financial data such as stock prices typically consists of numerical values that update frequently and may require high precision. Enhancing the algorithms to securely transmit different data types will expand the system’s usability beyond military applications.

  • Network limitations: the system assumes the availability of secure TCP/IP communication channels, which may not be reliable in all geographical contexts. In unreliable communication environments, data packets might arrive out of order, potentially compromising the integrity of the transmitted information. This limitation necessitates enhancements to the receiver-side algorithm to handle out-of-order image sequences, ensuring data integrity in less secure environments. Additionally, by exploring alternative secure communication methods, the system’s reliability can be further enhanced.

  • Storage space limitations: The system assumes that sufficient storage space is available at both the sender side and all nodes used during communication for the stego images and key image, which may not always be the case. To address this limitation, future improvements could focus on optimising storage utilisation through image compression or choosing tunable parameters in the embedding process, like the image sizes and the pixel count, to suit the available storage size. This will also enhance the system’s scalability and operational efficiency.

Scope for future enhancement

Considering the strengths and weaknesses of the system, several promising avenues for future exploration are evident as follows:

  • Adapting the system for secure transmission of sensitive medical records, financial transactions and IoT sensor data to broaden its applicability.

  • Enhancing the system to operate seamlessly across different operating systems and communication protocols. This will also increase its compatibility and usability.

  • Exploring how the system can integrate with emerging technologies like 6G networks and blockchain for enhanced performance and multi-layered security.

  • Identifying or curating a large database of carefully selected cover images that can support the practical implementation of the system without external support.

  • Conducting real-world practical testing and gathering expert feedback can further validate the system’s performance, reliability and security, and enhance its overall credibility.