Analysis of a p-persistent CSMA packetized cellular network with capture phenomena

A p-persistent CSMA protocol suitable for packetized mobile radiotelephone cellular networks with capture and a number of mobile terminals transmitting to a common receiver is considered. In a micro-cellular environment, messages from different transmitting terminals will suffer different attenuations (path loss and fading phenomena), yielding different levels of energy at the receiver. Thus, one out of 'i' attempting transmission packets will be successfully transmitted with some capture probability fi. Assuming that the receiver is within the range and in line-of-sight of all the mobile users in the cell, the capture probabilities can be determined by means of simulation, and a new probabilistic Markov model for the above network is introduced and analysed using the slot property of the channel Under moderate packet length size the results obtained in respect of the throughput-delay performance and stability under heavy traffic conditions are promising compared with those of the corresponding conventional p-persistent CSMA schemes. Consider a finite population of N mobile users sharing one channel for transmission to one receiver (base station). The channel is slotted with slot duration equal to T seconds. Any effect due to the propagation delay is assumed negligible. Each user generates packets of fixed length equal to 0 > 1 slots. Thus, the packet transmission time if OT seconds. Each user stores and attempts to transmit at most one packet for each time slot. A user without a packet in his buffer is said to be free; if he has a packet he is said to be ready. We assume a synchronized structure. Each user generates (or does not generate) a packet in a time slot, independently of other users, with probability cr (or 1-c7). Packet arrivals occur only from free users at the kT instants ;k-1,2 ..... and their number is distributed