Predicted Call and Residual Lifetime Based Channel Allocation Model for Primary User Equivalent QoS in Cognitive Radio Cellular Network

Abstract

Primary network channels follow binary on-off states with random time duration. Primary User (PU) traffic is observed each hour for channel allocation to secondary user (SU). As per available research works, on placement of SU call request at an instant, the channel allocation processor has to input (a) hourly call arrival rate (λ) of available channels till preceding hour to predict λ for current hour using SARIMA method, (b) Average call holding time in fraction of an hour from channel occupancy statistics and calculate blocking probability of different channels to offer to SU. Further, some optimistic research works excludes busy channels at the instant of SU call offer and selects some particular free channel based on prediction of ‘off period lifetime’. All the calculations are based on hourly traffic measurement where as call holding time is in minutes. The allocation of specific PU channel to SU cannot guarantee reliable Quality of Service (QoS). In present paper, PU traffic has been observed each minute for finer analysis. Minute-wise channel occupancy traffic is bumpy in nature, hence, present paper predicts λ using Holt Winters method. Also, at the instant of SU channel request, the channel allocation processor inputs all PU channel status minute-wise, calculates actual mean residual lifetime in minutes for each vacant channel and selects the channel with highest predicted free time. A simulation program runs on data collected from mobile switch of cellular network which creates pseudo-live environment for channel allocation. The present work has compared the MRL method with the other researchers using probabilistic method of channel allocation and MRL method has established as more accurate. The obtained result shows that QoS obtainable to SU is equivalent to PU even during busy hours.