Efficient SINR Queries for CSMA/CA Simulation

We propose an efficient simulation technique for the CSMA medium access protocol. It is based on the well-established model of using signal to interference and noise ratio (SINR), which is very accurate and allows for highly realistic predictions of collisions. However, traditional implementations require O(n) time to compute interference values. We evaluate how to speed this up by using efficient data structures such as k-d-trees or geometric hash tables in our implementation. There are different levels of accuracy, some of which allow for O(√n + a) lookups, where a is the number of nearby senders. We demonstrate the achievable speedup and discuss accuracy tradeoffs for different settings using the Shawn simulator.