Simulation of film transmission path based on ant colony optimization algorithm

Abstract

At present, the development and popularization of network technology has continuously innovated traditional film and television communication methods, and network film and television communication paths have become the new favorite of the public. Although various video apps for playing movies and TV series are emerging on the market, these apps are more sensitive to the terminal’s broadband, delay, delay jitter, and packet loss rate. Once the load is too heavy, it will seriously affect the running rate of the application, causing network congestion, thereby affecting the spread of movies and television. Based on this, this paper proposes an ant colony optimization algorithm (ACA) to conduct simulation experiments on film and television transmission paths, and on this basis, optimize and upgrade the path to improve the speed and scope of network film and television transmission. The research results show that the effect of the optimized ACA is significantly better than the basic ACA. The optimization effect of the propagation path is better.

Introduction

Background and meaning

With the rapid development of modernization, traditional film and television communication methods have been gradually eliminated. New film and television communication paths such as mobile phones, tablet computers, and Internet TV have gradually emerged among the public and quickly become a craze. Due to the limitation of multimedia video software itself, there are certain limitations in dissemination. In addition, the current Internet is developing rapidly, the quantity of nodes has increased exponentially, and the capacity and load of network transmission are gradually expanding, which poses new challenges to network quality of service (QoS).

ACA is a new type of bionic swarm intelligence algorithm. It is a new concept based on the foraging behavior and group characteristics of ants in nature. It is mainly used to study and calculate path planning and optimization problems. ACA has the characteristics of parallelism, positive feedback, high solution accuracy, and has strong robustness and global optimization ability. Aiming at the slow speed of network film and television transmission, the ACA can be better solved.

With the continuous development of network technology, software-defined networks and related technologies have gradually become an important direction for optimizing network resource allocation. Software-defined multimedia Internet of Things has significantly improved the efficiency of multimedia data transmission through centralized control and dynamic resource allocation. The core idea of these technologies is to achieve dynamic allocation and path optimization of network resources through software and virtualization, which has certain commonality with the film and television transmission path optimization method based on ant colony optimization algorithm proposed in this paper. In particular, the research on efficient resource allocation in software-defined Internet of Vehicles provides an important reference for this paper to optimize film and television transmission paths in complex network environments.

In the era of rapid technological advancement, the Internet of Things (IoT) has revolutionized multimedia transmission by enabling data collection and processing at the network edge, significantly enhancing the efficiency and responsiveness of multimedia delivery. In recent years, Green Cloud Multimedia Networking, as a key direction to improve energy efficiency and sustainability, has shown good potential in multimedia data processing and distribution. Green Cloud Multimedia Networking applies advanced technologies such as NFV (Network Functions Virtualization) and SDN (Software Defined Networking) to optimize the allocation and management of resources in cloud computing platforms and network facilities (especially data centers and WANs) to reduce significantly energy consumption in the distribution of multimedia content (such as cloud services, video streaming), so as to achieve the goal of environmentally friendly and energy-saving sustainable operations.

The integration of Software-Defined Internet of Things (SDIoT) further optimizes bandwidth allocation and device management, providing a dynamic framework for multimedia content distribution. Quality of Service (QoS) considerations are critical in SDIoT frameworks, ensuring reliable and high-quality multimedia streaming even under varying network conditions. The concept of Software-Defined Internet of Multimedia Things (SDM-IoT) extends these principles specifically to multimedia applications, offering enhanced flexibility and scalability in transmission. Furthermore, in the context of vehicular networks, Software-Defined Internet of Vehicles (SDIoV) plays a pivotal role in facilitating real-time multimedia streaming for both entertainment and informational purposes, addressing the unique challenges of mobility and latency. These advancements highlight the broader applicability and contemporary relevance of our research on ant colony optimization algorithms for film and television transmission paths, as they collectively aim to improve network efficiency and user experience in multimedia communication.

Related work

ACA, as a kind of AI bionic calculation method, has been applied in all aspects of life. Wu, Lei and Tian, Xue have conducted research on ACA regarding the drawbacks of pipeline routing design (PRD). To effectively solve the problems of oil and gas processing system and semi-submersible production platform, they proposed an optimized ACA. First, to keep the PRD problem simple, they constructed a new framework of the mathematical model based on actual constraints and rules, then an improved ACA is introduced, which combines one function and two mechanisms, namely heuristic function, mutation mechanism and dynamic parameter mechanism. They conducted experiments with a set of specific examples, and the experiments data showed that the overall performance of IACO is better than the two variants of ACO, particularly regarding the speed of convergence and the diversity range of different groups. Finally, they applied the mathematical model to actual engineering and solved the PRD problem of the semi-submersible production platform oil and gas processing system¹. In addition to pipeline routing, ACA is also used to solve the frequency problem of electromagnetic waves. Danny Z. Zhu and Pingjuan L. Werner said that frequency selective surface is widely used in the fields of electromagnetic wave and spatial filtering, including antenna, polarizer, radome and intelligent building. The traditional frequency selective surface has developed for a long time, and great changes have taken place in this process, from plane shape and pattern to multi-layer small design, and the stable response degree of the filter has reached 50. But at the same time, they pointed out that there is currently no powerful design tool to take advantage of this manufacturing method. Based on this, they introduced a lazy ant colony optimization algorithm featuring as multi-objective, which has adaptive characteristics. This algorithm is mainly applicable to the polarization design of three-dimensional frequency selective surface. The experimental results show that this algorithm can construct several lattice geometric structures, which are innovative and non-intuitive, deviation from the center frequency of less than 1%- and − 10-dB suppression and 6% 3 dB transmission bandwidth between − 12 is for the case where the maximum incident angle is 80° in TE and TM polarization². Although the ACA is widely used, there are few research on its spread in film and television. As an important way of people’s leisure and entertainment, film and television are indispensable and extremely important to the lives of the public. Therefore, the final research purpose of this paper is to explore and discuss the advantages and disadvantages of ACA in the optimization problem of film and television transmission path and hopes to find a solution to the problem of slow speed and narrow scope of current network film and television transmission, thereby broadening the path of network terminal film and television transmission and improving network service quality.

In the field of resource allocation, dynamic resource allocation strategies such as “Dynamic Offloading in Mobile Edge Computing with Traffic-Aware Network Slicing and Adaptive TD3 Strategy” provide new ideas. By combining network slicing and reinforcement learning strategies, the resource utilization in the edge computing environment is significantly improved. In addition, “Multi-objective Constellation Optimization and Dynamic Link Utilization for Sustainable Information Delivery Using PD-NOMA Deep Reinforcement Learning” proposes a multi-objective optimization method based on deep reinforcement learning, which provides an effective solution for sustainable information transmission. These studies provide important theoretical support and practical references for the film and television transmission path optimization method based on the ant colony optimization algorithm proposed in this paper^3,4,5.

Innovations in this article

This article has some innovations, as shown below: (1) Film and television entertainment is an important part of people’s lives. Studying film and television communication paths and broadening the network channels of film and television communication are essential for improving the quality of network services and satisfying more and higher demands of people. It is of great significance; (2) ACA is a new type of bionic artificial intelligence algorithm, which has strong robustness and path optimization ability. It plays a great role in solving network congestion and improving broadband communication and processing capabilities. The ACA is used to simulate the film and television propagation path, which will be of great help in optimizing the path; (3) Aiming at the slow convergence speed of the ACA, this paper proposes an ant colony optimization algorithm that combines multicast routing. The optimized ACA is suitable for the combination of larger quantity of nodes, because its network overhead is much smaller than the basic ACA.

ACA and film and TV transmission path

ACA

(1) Algorithm principle.

ACA, also known as ant algorithm, is a bionic artificial intelligence algorithm that simulates the group foraging behavior of ants in nature. It was proposed in the 1990s and was inspired by the behavior of ants finding paths in the collective foraging process⁶. ACA is a new type of general heuristic method that can be used to solve combinatorial optimization problems. It is a probabilistic technology that can find optimal paths in the graph. Ants are social animals, like humans, with a clear division of labor. Although they have no vision, they can use the pheromone in the environment to quickly distinguish the direction and find the path⁷. The basic principle is while foraging, the ant can leave a pheromone in the place it passes and transmit information with other ants. At the same time, during the movement process, the ant will forage according to the path with the pheromone. Activities, the more ants that walk on this path, the greater the pheromone concentration, and therefore the greater the probability that latecomers will choose this path, and the collective ant colony’s foraging behavior formed over time will show most ants walk on the path with the shortest distance between food and ant nest, which is the optimal path^{8]– [9}.

(2) Features of ACA.

The characteristics of ACA are as follows:

First, self-organization. When ant colonies are foraging, latecomers rely solely on pheromones to find foraging paths. Everyone cooperates with each other, divides labor separately, and completes work tasks collectively. The external interference is very small. They gradually find the optimal path in cooperation. The initial value is less dependent¹⁰.

Second, positive feedback. The ant colony relies on the ant’s pheromone to find the optimal path. The ant walking in front leaves the pheromone, and the following ants finds the optimal path with the accumulated pheromone. This process is a positive feedback process. In the first process of ACA, the pheromone of each path is the same, and when the external environment changes, the pheromone on each path also changes accordingly. At this time, there is a distinction between the pros and cons of the path, and the accumulation of pheromone more paths will have more ants, and paths with fewer pheromones will have fewer ants. This is positive feedback¹¹.

Third, distribution. Each ant in the ant colony has a clear division of labor, and everyone has independent behavioral capabilities. The individual solution can be completed independently regardless of whether it is in the same environment or in different environments, and when an individual’s solution process cannot be completed. Will not affect the overall solution, which makes the ACA has a good global search ability¹².

Fourth, scalability. The communication between the ants in the ant colony relies on the same pheromone to transmit information. This substance is universal. If you have one ant’s pheromone, you can attract all ants, which makes the ACA very good. Scalability^13,14.

Fifth, stability. Since the ant colony relies on pheromone to form the optimal path during the foraging process, the algorithm no longer needs to intervene in the calculation and search process, and only needs to set simple parameters, which also makes it easy to integrate with other algorithms to form new optimization algorithm^15,16.

(3) ACA steps.

Suppose there is $m$ant, ${l_{ij}}\left( {i,j=1,2, \ldots \ldots ,n} \right)$ represents the distance range between city $i$ and city $j$, and ${t_0}$ refers to the amount of pheromone left on route$\left( {i,j} \right)$ at time $t$.

Step 1: Initialize the parameters. Let $t=0$and the quantity of cycles ${N_a}=0$, set the maximum quantity of cycles to ${N_a}\hbox{max}$, place $m$ ants on $n$cities, and set the distance$\left( {i,j} \right)$ between every two cities to be initialized, which is${t_0}=const$, where$const$is a constant, and the initial time Point$\Delta {t_{ij}}\left( 0 \right)=0$, where $\Delta t_{{ij}}^{k}\left( t \right)$ refers to the total quantity of information left by the $k$ ant on the circular path.

The second step: open loop, let ${N_a}+1$.

The third step: ant index quantity $k=1$.

Step 4: Let the number of ants be $k+1$.

Step 5: According to the basic formula of transition probability, the transition probability of each ant is calculated, and selects city 4 to move forward. The transition probability formula is as (1)

$$P_{{ij}}^{k}\left( t \right)=\left\{ {\begin{array}{*{20}{c}} {\frac{{{{\left[ {{t_{ij}}\left( t \right)} \right]}^b}{{\left[ {{\eta _{ij}}\left( t \right)} \right]}^\beta }}}{{\sum\limits_{{s \in allowe{d_k}}} {{{\left[ {{t_{is}}\left( t \right)} \right]}^b}{{\left[ {{\eta _{is}}\left( t \right)} \right]}^\beta }} }},j \in allowe{d_k}} \\ {{\text{ }}0{\text{ ,otherwise}}} \end{array}} \right.$$

(1)

Among them,$allowe{d_k}$ represents the city that the $k$ ant can choose in the next step, and $b$refers to the factors inspired by information. The greater its value, the greater the probability that the ant will choose this transfer path. $\beta$represents the expected heuristic factor, and$\eta {}_{{ij}}\left( t \right)$ refers to the heuristic function. The circulation formula can be expressed as ${\eta _{ij}}\left( t \right)=\frac{1}{{{l_{ij}}}}$. Considering the high sensitivity of film and television data to delay and bandwidth, this paper increases the α value to enhance the algorithm’s reliance on historical paths, while reducing the ρ value to maintain the diversity of path exploration.

Step 6: Select the taboo table, move the ants to a new city, and move the city to the taboo table.

Step 7: The interaction between pheromone heuristic factors and expectation heuristic factors in the ant colony algorithm. The pheromone heuristic factor reflects the accumulation of pheromones on the path, guiding the ants to choose the optimal path from historical experience; while the expectation heuristic factor guides the ants to explore shorter paths based on the visibility and intuitive estimation of the path.

Step 8: If the city in the loop is not finished, which is$k<m$, then return to the fourth step for re operation, otherwise proceed to the next step.

Step 9: Use formula (2) the amount of information on each transfer path can be updated in time

$${t_{ij}}\left( {t+n} \right)=\left( {1 - \rho } \right){t_{ij}}\left( t \right)+\Delta {t_{ij}}\left( t \right)$$

(2)

Step 10: When the quantity of loops is${N_a} \geqslant {N_a}\hbox{max}$, the loop ends, otherwise it returns to the second step^17,18.

Ant colony optimization algorithm

(1) Multicast routing technology.

Multicast is a communication mode. A host sends a task, and all hosts that join the same group can receive the task. It realizes a point-to-multipoint network connection between the sender and the receiver, like ours group chat function in social software. Multicast improves the efficiency of data transmission and reduces the possibility of network congestion^19,20. Multicast technology is one of the three major methods of data transmission on IP networks. In addition to multicast, there are unicast and broadcast. Unicast is the point-to-point transmission between hosts, and broadcast is the delivery of a data packet from the host network to the hosts in each sub-network. Both unicast and broadcast have the problem of low efficiency and have certain limitations for the dissemination of data and information, while multicast only needs to send information once, and each host user network can receive the message, and the information can be quickly transmitted^21,22.

(2) Multicast routing algorithm.

The problem of multicast routing is mainly about the quality of network service. The algorithm aims to reduce the occurrence of network congestion by reducing the delay rate and packet loss rate of the network. The multicast routing algorithm is based on the network topology and the link state to establish a structure to realize the optimization target of the objective function under the constraint conditions and minimize the network cost. This constraint is not only the network service quality requirement, but also the multicast routing algorithm requirement, including delay, delay jitter, good packet loss rate and hop count. Here we use a heuristic algorithm and use spanning tree to calculate the multicast²³.

Suppose multicast spanning tree$T\left( {s,m} \right)$ is a spanning subtree of graph G, and this subtree covers the source and destination nodes, $m=\left| M \right|$is the quantity of multicast destination nodes, $M$ is defined as the destination node set, and$\left\{ {\left\{ s \right\}UM} \right\}$ is the example of multicast group set, ${P_t}\left( {s,d} \right)$ is a single path from the source node $s$to a destination node in the combination of multicast spanning tree. $T\left( {s,M} \right)$ is the comprehensive cost of all paths on the tree, then

$$C\left( {T\left( {s,M} \right)} \right)=\sum\limits_{{e \in T\left( {s,M} \right)}} {C\left( e \right)}$$

(3)

The total delay of path ${P_t}\left( {s,d} \right)$ refers to the quantity of the delays on the transfer path of ants

$$D\left( {{P_t}\left( {s,M} \right)} \right)=\sum\limits_{{e \in {P_t}\left( {s,d} \right)}} {D\left( e \right)}$$

(4)

The bottleneck bandwidth of path ${P_t}\left( {s,d} \right)$ refers to the value of the smallest available bandwidth on the path

$$B\left( {{P_t}\left( {s,d} \right)} \right)=\hbox{min} \left\{ {B\left( e \right),e \in {P_t}\left( {s,d} \right)} \right\}$$

(5)

The accumulated packet loss rate of path ${P_t}\left( {s,d} \right)$ refers to the product of the data arrival rate on the path

$$L\left( {{P_t}\left( {s,d} \right)} \right)=1 - \prod\limits_{{e \in {P_t}\left( {s,d} \right)}} {\left( {1 - N\left( e \right)} \right)}$$

(6)

The basic definition of the delay jitter of the spanning tree is the average value of the delay difference on the path from the source node to each target node, and its formula is expressed as follows:

$$DJ\left( {T\left( {s,M} \right)} \right)=\sqrt {\sum\limits_{{d \in M}} {{{\left( {D\left( {{P_t}\left( {s,d} \right) - delay - avg} \right)} \right)}^2}} }$$

(7)

Then the multicast routing problem that satisfies the network quality of service can be expressed as

$$MinC\left( {T\left( {s,M} \right)} \right)=\left\{ {\begin{array}{*{20}{c}} {D\left( {{P_t}\left( {s,d} \right)} \right) \leqslant \Delta d}&{,\forall d \in M} \\ {B\left( {{P_t}\left( {s,d} \right)} \right) \geqslant {B_d}}&{,\forall d \in M} \\ {L\left( {{P_t}\left( {s,d} \right)} \right) \leqslant {L_d}}&{,\forall d \in M} \\ {DJ\left( {T\left( {s,M} \right)} \right) \leqslant D{J_d}}&{,\forall d \in M} \end{array}} \right.$$

(8)

Latency and jitter are measured in milliseconds (ms), loss rate is measured in percentage (%), and bandwidth is measured in bits per second (bps).

(3) Improved ACA.

Combined with multicast routing technology, the basic ACA is optimized and improved. The idea is to add${\raise0.7ex\hbox{$e$} \!\mathord{\left/ {\vphantom {e {{T^{bs}}}}}\right.\kern-0pt}\!\lower0.7ex\hbox{${{T^{bs}}}$}}$pheromone to each side of the optimal path, where $e$ is a parameter that determines the weight of the optimal path. The pheromone update strategy is

$$\Delta {t_{ij}}\left( t \right)=\sum\limits_{{k=1}}^{m} {\Delta t_{{ij}}^{t}} \left( t \right)$$

(10)

$$\Delta t_{{ij}}^{k}\left( t \right)=\left\{ {\begin{array}{*{20}{c}} {\frac{Q}{{{L_k}}}} \\ 0 \end{array}} \right.$$

(11)

$$\Delta t_{{ij}}^{{bs}}\left( t \right)=\left\{ {\begin{array}{*{20}{c}} {\frac{Q}{{{L_k}}}} \\ 0 \end{array}} \right.$$

(12)

The optimized ACA can make more use of the current better solution, and it only adds pheromone on the edge to which the global optimal solution belongs. Whenever the ant moves from one point to another, the information on the edge element will appropriately reduce the path construction and greatly optimize the path structure.

Film and television transmission path method

Traditional film and television broadcasts are mainly cable TV, wireless broadcasting, and movie theaters. However, with the development of Internet technology, terminal devices such as smart phones, tablet computers, and Internet TV have emerged in large quantities, and the emergence of various video applications has further promoted the network. The development of film and television transmission path. In the network information technology environment, the path of film and television transmission is mainly based on the network supported by computer technology. A variety of mainstream video software, such as Tencent Video, iQiyi, Youku, etc., are used for network broadband, delay, delay jitter, Both the packet loss rate and network costs are more sensitive and require certain network service quality assurance. Moreover, most of the movies and TV dramas occupy a large amount and require high network operation speeds. Once the load is too large, network congestion will easily occur, which will affect the transmission of movies and TV. Therefore, it is very important to optimize the network transmission path of movies and TV. To this end, we use ACA and multicast routing technology to simulate the path of film and television propagation, increase the network operating speed from the source, and reduce network costs. The specific simulation experiment will be introduced in the next chapter.

Simulation experiment of film and television propagation path based on ant colony optimization algorithm

Description of the simulation system

In the algorithm simulation experiment, the search environment automatically created in the background is a grid with a length and width of 20 × 20. The starting point of the path object is the left grid, and the end point is the grid at the lower right corner. In the foreground, a path area is designed using OpenGL. The black net in this area represents obstacles, and the white net is a feasible path. To verify the feasibility and effectiveness of the algorithm, simulations were carried out on the path of film and television propagation.

Generate a random network

In the simulation experiment, if a specific network is used to simulate the algorithm, it is not convincing, because this network is not universal, so we choose a randomly generated network to simulate the film and television propagation path. The most important step in generating a random network is to generate a random network topology, and this network should be very similar to the actual network. Here we use Salama’s network topology generation algorithm, the algorithm steps are as follows:

Step 1: Calculate the maximum distance $d$ between any two nodes.

Step 2: Connect the first node with any two random nodes according to probability.

Step 3: Connect all isolated nodes with non-isolated nodes according to probability$P\left( {u,v} \right)$;

Step 4: Connect the node with degree 1 to a random node according to probability $P\left( {u,v} \right)$;

Step 5: Repeat the above steps continuously until the degree of all nodes in the network is greater than or equal to 2.

Simulation modeling of film and television propagation path

With the development of modern information technology, the advent of the era of big data is profoundly affecting people’s lives and production methods. Under the new technology situation, online film and television transmission paths have gradually become the mainstream. Popular film and television apps such as Tencent, Youku, and iQiyi are currently main platforms and paths for the spread of major TV dramas in our country. Based on the high requirements of this kind of APP for network service quality, a simulation model of film and television propagation path is designed and constructed using ant colony optimization algorithm.

(1) Experimental parameter setting.

Set the pheromone parameter on the path to$\alpha =1$, the visibility parameter to$\beta =1$, the pheromone volatilization coefficient $\rho =0.5$, the maximum quantity of cycles of the ant colony ${N_a}\hbox{max} =300$, and the number of ants $M=50$.

(2) System development and operating environment.

To find the optimal path for calculation, this study uses the software of Dev-C + + 5.11 (https://sourceforge.net/projects/orwelldevcpp/) as the development environment. The choice is based on the following considerations: the integrated environment has a built-in MinGW compiler, supports the C + + 11 standard and has lightweight characteristics, which is suitable for the rapid implementation and debugging of small and medium-sized algorithms; at the same time, its simple graphical interface reduces the complexity of experimental variables.

The experiment was run on the Windows XP SP3 system, mainly because the operating system has good compatibility with the drivers of early hardware devices and can reproduce the benchmark test conditions of similar algorithms in the literature.

(3) Path planning and optimization.

Path planning is to find an optimal path from the start point to the end point, that is, the shortest path. Usually, the film and television transmission path must go through a series of processes such as uploading, user browsing, downloading or storage, and the resource occupies a large proportion, and the network transmission speed and storage requirements are also high. The key to optimizing the film and television transmission path is to improve the upload speed and efficiency reduce the intermediate flow of propagation and minimize the path of propagation. In the generated random network topology, to find the optimal transfer path, we generally choose to use the ant colony optimization algorithm to achieve this goal.

(4) Simulation modeling.

Combining multicast routing technology to optimize the ACA, and finally establish a film and television propagation path model, as shown in Fig. 1.

Fig. 1

Film and television propagation path simulation model.

In the simulation modeling of film and television transmission paths, in order to more realistically reflect the complexity of multimedia transmission environments, this paper introduces dynamic network parameters based on a 20 × 20 grid, including variable link delay, jitter, adaptive bit rate, and congestion control mechanism. These parameters provide a more challenging test environment for path optimization by simulating the uncertainty in the actual network. To correspond the behavior of the ant colony algorithm to the actual network behavior, this paper establishes a mapping mechanism between ant path selection and packet routing. The movement of ants in the grid is mapped to the transmission of data packets in the network, and the pheromone concentration corresponds to the link quality, so that the simulation results can more directly reflect the QoS indicators in the actual network. In combination with adaptive bit rate and caching strategy, this paper conducts a more comprehensive evaluation of the effect of multicast tree optimization.

Simulation experiment results

2015–2019 our country’s online video application software user scale and utilization rate trend

In the 21st century, network technology has developed rapidly. Therefore, people use online video apps to watch movies and TV series more and more frequently, and third-party video applications developed by major video websites have been downloaded by most mobile terminal users. Figure 2 shows the user scale and usage trend of our country’s online video application software from 2015 to 2019.

Fig. 2

The user scale and utilization rate trend of my country’s online video application software from 2015 to 2019.

As can be seen from Fig. 2, in recent years, the transmission path of film and television has gradually become networked and digitized. From 2015 to 2019, the user scale and utilization rate of our country’s online video application software have generally increased, and the quantity of online video APP users nationwide reached 591.67 million in 2019. People, its utilization rate reached 64.89%. This shows that the transmission path of online film and television has become the mainstream method of film and television transmission.

Influence of various parameters on the planning of film and television transmission paths

In the film and television propagation path simulation model designed and constructed in this paper, the main parameters are the number of ants, the pheromone heuristic factor, the expected heuristic factor, the pheromone volatilization factor, and the pheromone concentration. The above parameters will have a direct influence on the overall performance and effect of the algorithm, and then affect the final optimizing the path. In the simulation experiment, we recorded the values of the relevant parameters and plotted them into Table 1 and compared and analyzed the relationship between the parameters.

Table 1 The effect of the number of ants on the path length and running time.

It can be seen from Table 1 that the optimal path length and average path length of ants will increase as the total number of ants rises.

Table 2 The influence of pheromone heuristic factors and expected heuristics factor on the results of the algorithm.

It can be seen from Table 2 that in the combination of the pheromone heuristic factor and the expected heuristic factor, the average path length, and the optimal path length of the second group are both optimal, and when the pheromone heuristic factor is greater than the expected heuristic factor, the ant group algorithm performs best.

Table 3 The influence of information volatilization factors on the results of the algorithm.

It can be seen from Table 3 that the optimal path length and average path length of the ants will roughly increase with the increase of the volatilization factor.

Table 4 Effect of pheromone concentration on path length and running time.

According to the data in Table 4 above, it can be clearly concluded that when the concentration of pheromone gradually increases, the moving time of ants gradually shortens, and the average path length is also gradually decreasing, while the optimal path length is basically unchanged. According to the data in Tables 4 and 30 groups show the best indicators. The reason may be that it is the closest value to the largest pheromone parameter in the simulation model.

To further evaluate the interaction between parameters, this paper designs a simplified orthogonal experiment to simulate the impact of the combination of α, β and ρ on the average path length of the algorithm. The results are shown in Table 5.

Table 5 Parameter sensitivity analysis.

Overall, the path length fluctuates significantly between different combinations, indicating that there is a complex interaction between the parameters. The essence of this phenomenon reflects the control mechanism of the parameters on the search behavior in the ACA algorithm. α controls the ant’s dependence on historical pheromones. The larger the α value, the more likely the ant is to repeat the existing path; β reflects the importance attached to the heuristic function. The larger the value, the more preference for short paths; and ρ represents the pheromone volatilization speed, which directly affects the memory update mechanism of path exploration. When α and β are set reasonably, such as the combination of Group 5, the algorithm can better strike a balance between using historical experience and finding new paths, while a moderate ρ value keeps the pheromone concentration properly, avoiding forgetting too quickly and not falling into the local optimum. On the contrary, when α is too high and ρ is too low, such as Group 8, due to excessive reliance on existing paths and slow pheromone volatilization, the algorithm will fall into early convergence and it will be difficult to jump out of the suboptimal solution, affecting the optimization effect of the final propagation path.

Performance comparison between ACA and ant colony optimization algorithm

(1) Convergence performance analysis.

In the statistical simulation experiment, the cycle times of the two algorithms are tested and the cost of the multicast spanning tree changes as the cycle times increase. In the simulation experiment, the quantity of network nodes for these two algorithms is 50, and the quantity of target nodes is 5. Figure 3 shows the convergence performance of these two algorithms.

Fig. 3

Convergence performance analysis of the two algorithms.

According to the data of Fig. 3, we can conclude that in the initial stage of the cycle, the convergence speed of the basic ACA is obviously faster than that of the ant colony optimization algorithm, but when the quantity of cycles reaches 50, it basically no longer decreases, and its convergence is optimal at this time; The optimization algorithm starts to take effect after the quantity of cycles reaches 70, and the convergence period is shorter than the basic ACA. This is because the ant colony optimization algorithm combines multicast routing technology, which greatly reduces the quantity of network nodes and enhances the randomness of the path search, The improved ant colony optimization algorithm has a relatively slower convergence speed than the basic ACA, but the improved ACA is more expensive than the basic ACA on the optimal multicast growth tree.

(2) Network cost performance analysis.

Set the quantity of network nodes unchanged at 50, and the quantity of network destination nodes from 5 to 13. For the optimal multicast growth tree of each destination set size, the quantity of repetitions is taken as the average value of each optimal multicast growth tree. The optimal multicast growth tree cost for the situation. The result is shown in Fig. 4.

Fig. 4

The impact of the quantity of network destination nodes on network costs.

According to the data in Fig. 4 above, we can clearly see that when the quantity of target nodes in the network increases, the network cost of these two algorithms also increases. When the quantity of network nodes is small, the basic ACA and the improved ant colony optimization algorithm will increase. The network cost of the swarm optimization algorithm is not much different, and as the quantity of destination nodes gradually rises, the network cost of the two algorithms is significantly different. The network cost of the improved ant colony optimization algorithm is less than that of the basic ACA, which shows that this article proposed ant colony optimization algorithm has better performance and is more suitable for the situation where the quantity of network destination nodes is large. It is the best algorithm for path optimization.

Set the quantity of network destination nodes unchanged and analyze the performance of the network cost from 10 to 50. For optimal multicast growth trees with different quantities of network nodes, the quantity of repetitions takes the average value of each optimal multicast growth tree as the optimal multicast growth tree cost in this case. The result is shown in Fig. 5.

Fig. 5

The impact of the quantity of network nodes on network costs.

As shown in Fig. 5, as the quantity of network nodes rises, the cost of the optimal multicast growth tree, that is, the network cost, also increases; when the quantity of network nodes is less than 40, the basic ACA and the improved ant colony optimization algorithm. There is no obvious difference in the network cost of the ACA, and when the quantity of network nodes is greater than 40, the network cost of the two algorithms obviously becomes larger; when the quantity of network nodes is 50, the network cost of the basic ACA reaches 1500, while currently, the network cost of the improved ant colony optimization algorithm is 1100. This shows that as the quantity of network nodes increases, the effect of the optimized ACA is more and more different from the basic ACA.

To further evaluate the computational scalability of the proposed algorithm in larger and more realistic network scenarios, a set of controlled simulation experiments were conducted to compare the execution time of basic ACA and improved ACA at different network scales. Both algorithms were tested with increasing numbers of network nodes (n = 100, 200, 500) while keeping consistent ant population, pheromone dynamics, and heuristic influence parameter settings. The experiments aim to examine how the integration of multicast routing and optimized pheromone update strategy in improved ACA affects the runtime performance when the network complexity increases. Table 5 lists the results of the runtime benchmarks.

Table 6 Comparison of runtime of basic ACA and improved ACA at different network scales.

According to the data in Table 6 above, as the number of network nodes increases, the gap between the basic ACA and the improved ACA in running time gradually widens. When there are 100 nodes, the average running time of the basic ACA is 954ms, while that of the improved ACA is 712ms, a reduction of 25.4%; when the number of nodes is expanded to 200, the running times of the two are 1987ms and 1398ms respectively, and the optimization range is increased to 29.7%; in a large-scale network of 500 nodes, the running time of the basic ACA rises to 3625ms, while the improved ACA is controlled at 2270ms, with the most obvious improvement effect, saving 37.4% of the running time.

From the perspective of algorithm mechanism, this performance difference mainly stems from the optimization of the improved ACA in the path search strategy. The basic ACA traverses and evaluates all possible paths in each round, and its computational complexity increases quadratically as the network scale expands. The improved ACA effectively compresses the search space while maintaining the diversity of the solution space by introducing a multicast routing structure and an enhanced pheromone mechanism for the global optimal path. At the same time, the improved version also uses path pruning and expectation guidance strategies to make the algorithm tend to the potential optimal solution area in the initial stage. Especially in large-scale scenarios such as 500 nodes, the basic ACA is prone to invalid search due to the lack of structured guidance, resulting in a sharp increase in running time. The improved ACA alleviates this problem and shows good scalability and engineering practical value.

Discussion

Although the improved ACA algorithm shows better performance in multicast routing optimization, it may have limitations in highly dynamic or unpredictable network conditions. For example, frequent changes in network topology or sudden drops in link quality may cause the algorithm to fail to adapt in time, affecting the path optimization effect. In addition, in the face of a large amount of burst traffic, the algorithm may cause short-term network congestion due to untimely adjustment of the path. This shows that although the improved ACA performs well in a stable network environment, it still needs to be further optimized in an extremely dynamic environment.

Compared with the traditional ACO algorithm, the improved ACA proposed in this paper shows obvious advantages in convergence speed and network cost. However, to fully evaluate the performance of the algorithm, it needs to be compared with more ACO variants and other multicast routing algorithms. In addition, some algorithms in existing research may perform better in specific scenarios, so the applicability and limitations of these algorithms need to be further explored.

The improved ACA in this paper is innovative in combining multicast routing technology, especially in pheromone update strategy and path selection mechanism. The core performance analysis of this paper shows that the improved ACA outperforms the basic ACA when the number of destination nodes increases. To make these results more convincing, future research should include more comparative benchmarks, such as other multicast routing protocols or commercial solutions. In addition, the performance analysis should be extended to more network indicators, such as throughput, delay variation, and packet loss rate, to fully evaluate the practical application effect of the algorithm.

The enhanced ACO demonstrates improved performance in optimizing film and television transmission paths. However, its effectiveness may be constrained in highly dynamic or unpredictable network environments. A detailed analysis of the algorithm’s computational complexity reveals that the time complexity is primarily O(m·n²), where m represents the number of ants and n denotes the number of network nodes. This complexity arises from the need to evaluate all potential paths and update pheromone levels accordingly. Future research should include empirical studies to assess the algorithm’s robustness under varying network conditions, such as network failures, topology changes, and fluctuations in bandwidth and delay. These assessments are crucial for validating the algorithm’s practical applicability in real-world multimedia transmission scenarios.

In the experiments of this paper, preliminary observations show that when the value of α is high, ants are more inclined to choose paths with high pheromone concentrations, which helps to speed up the convergence speed, but limits the diversity of path exploration to a certain extent; on the contrary, when the value of α is low, the ants’ dependence on pheromones is weakened, and they are more inclined to explore new paths, which is conducive to jumping out of the local optimal solution and finding a better path. The sensitivity analysis of the expected heuristic factor β shows that a larger β value makes ants more dependent on heuristic information and tend to choose shorter paths, which is conducive to improving the convergence speed of the algorithm but may cause the algorithm to fall into the local optimal solution too early; a smaller β value makes ants more random in path selection, increases the diversity of path exploration, and is conducive to finding a better solution. As for ρ, its value determines the persistence of pheromones. Too low ρ will cause the algorithm to quickly forget the historical path information, resulting in a slow convergence speed; while too high ρ will make the algorithm too dependent on the initial path selection, making it difficult to effectively explore new paths, thus falling into the local optimal solution. These parameters affect each other and jointly determine the exploration ability and convergence speed of the algorithm, which has a significant impact on the overall performance of the algorithm. Future research can further explore the mechanisms of these interactions by designing orthogonal experiments or applying multifactor analysis methods to provide more comprehensive theoretical support for the application of ant colony algorithms in complex network environments.

Conclusions

With the development of Internet technology, film and television communication paths are becoming more and more diversified. Networked and digital communication paths have become the mainstream of the contemporary era. The third-party video application software developed and launched by major video websites has further broadened the film and television communication. However, these applications have higher requirements for network broadband and service quality. Only by continuously improving network service quality, reducing network congestion, and optimizing the path of film and television transmission, can the spread of film and television culture be better promoted.

As a kind of bionic artificial intelligence algorithm, ACA shows strong superiority in planning internal path selection problems, but the basic ACA has the disadvantage of slow convergence speed in calculation. Therefore, this research proposes to combine multicast routing Technology, the basic ACA is optimized and greatly improved, and the improved ACA has obvious advantages in choosing the best path.

This research uses the ant colony optimization algorithm to implement the simulation experiment of the film and television propagation path. The final experimental consequences show that the number of ants, pheromone heuristic factor, expected heuristic factor, pheromone volatilization factor and pheromone concentration have a certain influence on the performance of the algorithm; In the performance comparison between the basic ACA and the ant colony optimization algorithm, we finally found that the effect of the optimized ACA is significantly better than the basic ACA, and it has a better effect on the optimization of the film and television transmission path.