| Popis: |
The distributed scheduling mode possesses numerous advantages over the centralized scheduling mode in scenarios where the topology changes rapidly, such as low operation and maintenance costs, high robustness and improved delay performance, and thus has broad application prospects in robot swarms, device-to-device networks, industrial sensor networks, etc. The distributed election mechanism can effectively reduce collisions and improve the success probability of random access, which is one of the most commonly used channel access methods in distributed networks. The size of election interval directly affects the probability of successful election and the network performance. Specifically, with the expansion of the network scale, smaller election intervals will lead to more election nodes and reduce the probability of successful elections. Although expanding the election interval can reduce collisions to a certain extent, the number of legal election nodes will increase, which reduces the probability of successful elections. Therefore, the choice of the election interval needs to be explored in detail. This paper analyzes the relationship between the election interval and the number of effective election nodes, and establishes the analytical model of the election success probability, scheduling delay and effective throughput. In addition, the election interval size and the number of control slots are optimized to improve effective throughput and delay performance. Finally, the impact of the election interval on the network performance is verified by simulation, which provides guidance for parameter selection in realistic scenarios. |
