Abstract:Proposals for molecular communication networks as part of a future internet of bio-nano-things have become more intricate and the question of practical implementation is gaining more importance. One option is to apply detailed chemical modeling to capture more realistic effects of computing processes in biological systems. In this paper, we present ChemSICal, a detailed model for implementing the successive interference cancellation (SIC) algorithm for molecular multiple access in diffusion-based molecular communication networks as a chemical reaction network (CRN). We describe the structure of the model as a number of smaller reaction blocks, their speed controlled by reaction rate constants (RRCs). Deterministic and stochastic methods are utilized to first iteratively improve the choice of RRCs and subsequently investigate the performance of the model in terms of an error probability. We analyze the model's sensitivity to parameter changes and find that the analytically optimal values for the non-chemical model do not necessarily translate to the chemical domain. This necessitates careful optimization, especially of the RRCs, which are crucial for the successful operation of the ChemSICal system.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is: how to achieve multi - access (MA) in diffusion - based molecular communication networks (DBMC), especially non - orthogonal multi - access (NOMA), and successfully implement the successive interference cancellation (SIC) algorithm on this basis. Specifically, the authors propose a chemical reaction network (CRN) model named ChemSICal, which is used to implement the SIC algorithm through chemical reactions in a biological environment, thereby distinguishing the information sent by multiple transmitting nodes (TXs).
### Specific description of the main problem
1. **Multi - access (MA) problem**:
- In the scenario of Internet of Bio - Nano - Things (IoBNT) in the future Internet of Things, it is necessary to support the simultaneous transmission of multiple communication nodes.
- Several multi - access schemes are proposed, such as time - division multiple access (TDMA), molecular - division multiple access (MDMA), and non - orthogonal multi - access (NOMA). NOMA allows multiple nodes to use the same molecular type for communication simultaneously, but requires the implementation of the SIC algorithm at the receiving end (RX) to separate the signals of different nodes.
2. **Chemical reaction network (CRN) to implement the SIC algorithm**:
- Since the devices in the DBMC network are usually tiny biological structures with limited computing power (such as engineered cells or bacteria), general - purpose computing resources cannot be used.
- Therefore, a solution based on chemical reactions is required to implement the SIC algorithm in these restricted environments.
3. **Parameter optimization and performance evaluation**:
- The reaction rate constants (RRCs) are crucial to the system performance. The authors optimize the RRCs through deterministic and stochastic methods and evaluate the performance of the ChemSICal model.
- The sensitivity of the model to parameter changes is analyzed, and it is found that the theoretically optimal non - chemical model parameters are not necessarily applicable to the chemical domain, so re - optimization is required to adapt to chemical behavior.
### Solutions
- **ChemSICal model**:
- A detailed CRN model for implementing the SIC algorithm is proposed.
- The model consists of multiple small reaction modules, each of which is responsible for performing simple operations (such as comparison, conversion, approximate majority, etc.), and the reaction speed is controlled by adjusting the RRCs.
- **Optimization and evaluation**:
- The model is evaluated using deterministic and stochastic solvers (ODE and SSA), and it is found that the RRCs are the key parameters.
- Through preliminary sensitivity analysis, the detection threshold and RRCs are adjusted, significantly reducing the error probability.
### Conclusions
This research demonstrates the ability of the ChemSICal model to implement the SIC algorithm in the DBMC network and emphasizes the differences between chemical implementation and theoretical analysis. Future work will focus on further optimizing and expanding these concepts, for example, introducing timing mechanisms and chemical clocks to achieve faster and more controllable CRN execution.
### Formula summary
- The formula for the average number of received molecules:
\[
\lambda_i(t)=\frac{N_{\text{TX}} V_{\text{RX}}}{(4\pi D t)^{\frac{3}{2}}} \exp\left(-\frac{d_i^2}{4Dt}\right)
\]
- The probability mass function (PMF) of the received sample:
\[
p_{n_s}[n]=\frac{1}{4}\cdot\left(P_{\text{PMF}}(n; 0)+P_{\text{PMF}}(n; \lambda_2)+P_{\text{PMF}}(n; \lambda_1)+P_{\text{PMF}}(n; \lambda_1+\lambda_2)\right)
\]
- The probability mass function of the Poisson distribution:
\[
P_{\text{PMF}}(n; \lambda)=\frac{\lambda^n e^{-\lambda}}{n!}
\]
Through these methods, the paper effectively solves the problem of achieving efficient multi - access in a biological environment.