California Partners for Advanced Transportation Technology

This report investigates the performance of model-based estima-tion over multi-access networks and emphasizes on the estimationMSE (mean squared error) while using different channel access schemes:probabilistic (random access), deterministic (round-robin scheduling),and combined (grouped channel access). We propose a mathematicalframework for estimation over a simple multi-access MAC protocol,the slotted ALOHA network. Estimation MSE, its asymptotic be-havior and stability condition are derived for different channel accessmethods. Our quantitative discussion can provide guidelines to designthe communication logic for those vehicular control systems built ontop of multi-access networks.