Performance evaluation of R-ALOHA in distributed packet radio networks with hard real-time communications

The performance of reservation ALOHA (R-ALOHA) to support hard real-time communications in a distributed packet radio network is evaluated analytically and by simulation. The network requirements are that each node send an update packet to its communication partner at least every T seconds (called the deadline period) with high reliability. Two scenarios are considered: the worst case and the steady-state case. For the worst-case scenario, two Markovian models are used to investigate how fast the network nodes can recover the communications after a short period of catastrophic failure of the network (i.e., recover reservation). For the steady-state scenario, a Markov chain is formulated to study the trade-off between the key system parameters such as the channel data rate, the packet error probability, the deadline period, and the reliability requirement. One application of this network is for vehicle-to-vehicle communications in intelligent vehicle highway systems where a vehicle periodically sends its updated status (velocity, acceleration, etc.) to the following vehicle to improve safety and the vehicular traffic capacity.