Performance Analysis of Sequential Method for Handover in Cognitive Radio Systems

Powerful spectrum handover schemes enable cognitive radios (CRs) to use transmission opportunities in primary users’ channels appropriately. In this paper, we consider the cognitive access of primary channels by a secondary user. We evaluate the average detection time and the maximum achievable average throughput of the secondary user when the sequential method for hand-over (SMHO) is used. We assume that a prior knowledge of the primary users’ presence and absence probabilities are available. When investigating the maximum achievable throughput of the secondary user, we end into an optimization problem, in which the optimum value of sensing time must be selected. In our optimization problem, we take into account the spectrum hand over due to false detection of the primary user. We also propose a weighted based hand-over (WBHO) scheme in which the impacts of channels conditions and primary users’ presence probability are considered. This Spectrum handover scheme provides higher average throughput for the SU than the SMHO method. The tradeoff between the maximum achievable throughput and consumed energy is discussed, and finally an energy efficient optimization formulation for finding a proper sensing time is provided.

💡 Research Summary

This paper investigates spectrum handover strategies for a secondary user (SU) operating in a cognitive radio environment where multiple primary users (PUs) occupy licensed channels. Two approaches are examined: the Sequential Method for Handover (SMHO), which scans channels in a fixed numeric order, and a proposed Weighted-Based Handover (WBHO) scheme that orders channels according to a weight reflecting channel quality and PU arrival probability. The authors model PU activity with a two‑state (ON/OFF) Markov chain, derive the steady‑state idle probabilities, and adopt energy detection (ED) for spectrum sensing. Detection and false‑alarm probabilities are expressed via Q‑functions, linking the sensing time τ and decision threshold λ.

For SMHO, the probability that the SU transmits on channel k is q_k = P_fa·P_k,0 + P_d·P_k,1. The average number of handovers g_SMHO is obtained by summing the probabilities of consecutive busy detections before finding an idle channel, subject to a maximum handover count α limited by the slot duration T and the per‑hand‑over time τ_ho. The average sensing time becomes τ + g_SMHO·(τ+τ_ho). Using these quantities, the average normalized throughput R is derived as a product of a transmission‑efficiency term (depending on channel capacities under H0 and H1) and a time‑allocation term that accounts for sensing and handover overhead.

The optimization problem seeks τ and λ that maximize R while satisfying constraints on false‑alarm probability (P_fa ≤ P_max_fa), detection probability (P_d ≥ P_min_d), and τ < T. Because λ influences both q_k and the time‑allocation factor, the problem is inherently two‑dimensional; however, fixing P_d at its minimum allowable value reduces it to a one‑dimensional search over τ, with λ determined from the detection constraint. The authors show that decreasing λ increases q_k (more frequent handovers) and thus reduces the effective transmission time, creating a trade‑off that must be balanced.

The WBHO scheme introduces a weight w_i for each channel i, combining the channel’s average SNR, PU arrival probability, and possibly other quality metrics. Channels are sorted in descending order of w_i before sensing. This reordering reduces the expected number of handovers and the total sensing time, leading to higher throughput and lower energy consumption. Numerical simulations with five PU channels, a slot length of 10 ms, and a sampling frequency of 1 MHz demonstrate that WBHO achieves 15‑25 % higher average throughput than SMHO under identical false‑alarm constraints, while also consuming less energy.

The study highlights that in cognitive radio systems, the choice of sensing duration, decision threshold, and handover sequence critically impacts overall performance. SMHO offers simplicity but suffers from higher handover overhead, whereas WBHO leverages channel statistics to improve efficiency. The paper suggests future extensions such as collaborative sensing among multiple SUs, dynamic weight adaptation, and real‑time channel state estimation to further enhance spectrum access strategies.