TRAP: a high performance protocol for wireless local area networks

Abstract

The fixed number of available random addresses for randomly addressed polling (RAP) poses a significant problem in terms of scalability in case of many active stations. In such cases, the protocol's performance is significantly degraded. In this paper, we propose a TDMA-based randomly addressed polling (TRAP) protocol. The protocol employs a variable-length TDMA-based contention stage with the length based on the number of active stations. At the beginning of each polling cycle, the base station invites all active mobile stations to register their intention to transmit via transmission of a short pulse. The base station uses the aggregate received pulse in order to obtain an estimate of the number of contending stations and schedules the contention stage to comprise an adequate number of time slots for these stations to successfully register their intention to transmit. Then it transmits a READY message carrying the number of time slots P. Each mobile station calculates a random address in the interval [0… P−1], transmits its registration request in the respective time slot and then the base station polls according to the received random addresses. Simulation results are presented that reveal the superiority of TRAP against the RAP protocol in case of medium and high offered loads. Furthermore, the implementation of the proposed protocol is much simpler than that of CDMA-based versions of RAP, since no extra hardware is needed for the orthogonal reception of the random addresses.