Membrane Computing and Neural Processing Systems

The field of membrane computing and neural processing systems merges the principles of biological compartmentalisation with the computational paradigms of neural networks. Membrane computing, or P system theory, models the structure and dynamics of living cells through hierarchies of membranes that encapsulate multisets of objects and evolution rules. Neural processing systems extend this framework by integrating spiking neural P systems, in which information is encoded as discrete spikes exchanged between neuron-like compartments. This hybrid approach capitalises on the inherent parallelism, modularity and distributed nature of biological systems, offering new strategies for solving complex problems. Research has demonstrated that variants of spiking neural P systems can achieve Turing universality, simulate biochemical processes, and address real-world applications such as fault diagnosis in power grids. Recent developments have focused on enhancing computational efficiency, establishing formal verification methods and exploring hardware implementations, thereby reinforcing the global relevance of this discipline for both theoretical computer science and applied engineering.

Research from Nature Portfolio

Recent studies have investigated the computational power of spiking neural P systems with self-organisation, demonstrating that such networks can achieve Turing universality with a minimal number of neurons while preserving efficient parallel processing. These self-organising variants adapt their internal connectivity dynamically, enabling the simulation of any recursive function within a bounded resource envelope. This foundational work has clarified the theoretical limits of membrane-based neural models and opened pathways to more compact designs capable of universal computation under biologically plausible constraints.

Research from all publishers

A tutorial-style overview has synthesised the latest variants of spiking neural P systems, contrasting derivation modes, encoding schemes and performance metrics. This work provides practitioners with a roadmap to select appropriate models for specific computational tasks and outlines open challenges in achieving real-time efficiency. Complementary research has introduced sparse matrix-vector representations to accelerate the parallel simulation of spiking neural P systems on graphics processing units. By utilising compressed storage formats, the new algorithms reduce memory footprints and enhance computation throughput, making large-scale neural models feasible. In parallel, a comprehensive survey of fault-diagnosis applications has showcased the deployment of spiking neural P systems in monitoring power transmission networks. This account highlights automated methods for constructing diagnosis models, integrating fuzzy reasoning and rule-based spike communication to detect and localise faults with high accuracy and robustness against incomplete data.

Membrane Computing and Neural Processing Systems publication trend

The graph below shows the total number of articles in membrane computing and neural processing systems across all publications each year (not limited to Nature Index journals).

Technical terms

Membrane computing: A branch of natural computing that models computation through hierarchically arranged membranes with multiset rewriting rules.

P system: A formal computational model inspired by cellular structures, consisting of membranes, objects and evolution rules.

Spiking neural P system: A class of P systems in which neurons communicate via discrete electrical pulses or spikes according to spiking rules.

Turing universality: The capacity of a computational model to simulate any Turing machine and thus compute any computable function.

Sparse matrix representation: A storage scheme that records only non-zero elements of a matrix to reduce memory usage and speed up numerical operations.

References

1. On the Computational Power of Spiking Neural P Systems with Self-Organization. Scientific Reports (2016).
2. Spiking neural P systems: main ideas and results. Natural Computing (2022).
3. Spiking neural P systems: matrix representation and formal verification. Journal of Membrane Computing (2021).
4. A Review of Power System Fault Diagnosis with Spiking Neural P Systems. Applied Sciences (2021).
5. Simulation of Spiking Neural P Systems with Sparse Matrix-Vector Operations. Processes (2021).