Quadratic Greedy Algorithm for Multichannel Access in Cognitive Radio Network

Abstract

Spectrum handoff is one of the main issues in cognitive radio networks. Spectrum handoff occurs when the owner of the spectrum or the primary user reclaims a channel already used opportunistically by a secondary user. Such handoff becomes more frequent especially in case of multichannel access since that in a wider band, a PU is likely to reclaim a part of the band more quickly. This process has a negative impact on the performance of SU transmission in terms of delay and link maintenance. In this paper, the problem of minimizing the secondary user handoff process because of primary user activity is underlined. An analytical formulation based on knapsack problem is established and a new spectrum selection scheme is proposed. The main idea is to maximize jointly the holding time and the achievable throughput of the selected channels. The performance of the proposed access model is investigated through numerical simulations. KeywordsHandoff, Holding Time, Rate, Knapsack problem, heuristic, quadratic greedy algorithm. CONTEXT/MOTIVATIONSCognitive radio (CR) technology is regarded as a promising solution to the spectrum scarcity problem. The key idea of this paradigm is to allow unlicensed Secondary Users (SUs) to utilize opportunistically "spectrum holes" with the main condition to not interfere with the licensed Primary User (PU) transmission.However, due to the spectrum varying nature of CR networks, unlicensed users are required to perform spectrum handoff (HO), which refers to the process that when the current channel used by an SU is no longer available, the SU needs to pause its on-going transmission, vacate that channel, and determine a new available channel to continue the transmission.Clearly, many interruptions from the primary users specially in case of multichannel access will result in multiple handoffs, thereby affects negatively the performance of the SU transmission in term of delay and link maintenance. In fact, in normal conditions of licensed usage, channel switching can cause a packet loss ratio of 3% [1], this ratio becomes worse in CR usage because of the unpredictable behavior of PU. Besides, handoff generates additional delay since for each frequency switching a sensing phase has to take place, additionally to channel evacuation and link set up.Although reducing the HO due to PU activity is crucial for secondary transmission, limited studies consider this issue compared with other functionalities (spectrum sensing, spectrum management, and spectrum sharing) [2] of CR networks.Moreover, in the literature, most of research works regarding spectrum handoff focused on the physical-layer impacts without considering the effect in the MAC layer. Most of MAC protocol design for spectrum handoff falls into two categories based on the moment when SUs carry out spectrum handoffs. In the first category, SUs perform channel switching after detecting the reappearances of PUs, namely the reactive approach However, either for reactive or proactive HO process when SUs perform spectrum handoffs a well...

Greedy Based Proactive Spectrum Handoff Scheme for Cognitive Radio Systems

“…Due to the efficiency of greedy strategy, several studies have been done on spectrum handoff schemes implementing greedy algorithms, e.g. auction and greedy based spectrum allocation in multiple bands [1], quadratic greedy approach and handoff scheme based on knapsack model for multichannel selection and access [2], greedy and partially observable Markov decision process (POMDP) approaches for maximizing the throughput capacity [3]. Work [4] investigates and compares the performance of three greedy based schemes: P RE , SL(K), k t h-UCB1 for decentralized blind spectrum selection.…”