QoS-aware distributed spectrum sharing for heterogeneous wireless cognitive networks

Abstract

Ubiquitous wireless networking calls for efficient dynamic spectrum allocation (DSA) among heterogeneous users with diverse transmission types and bandwidth demands. To meet user-specific quality-of-service (QoS) requirements, the power and spectrum allocated to each user should lie inside a bounded region in order to be meaningful for the intended application. Most existing DSA methods aim at enhancing the total system utility. As such, spectrum wastage may arise when the system-wise optimal allocation falls outside individual users’ desired regions for QoS provisioning. The goal of this paper is to develop QoS-aware distributed DSA schemes using game-theoretic approach. We derive DSA solutions that respect QoS and avoid naively boosting or sacrificing some users’ utilities to maximize the network spectrum utilization. Specifically, we propose two game-based DSA algorithms: one resorts to proper scaling of the transmission power according to each user’s useful utility range, and the other embeds the QoS factor into the utility function used during gaming. To evaluate DSA schemes from a practical QoS perspective, we introduce two new metrics, namely “system useful utility” and “fraction of QoS-satisfied users”. Simulations confirm that the proposed DSA techniques outperform existing QoS-blind game models in terms of the spectrum sharing efficiency in heterogeneous networks. Convergence analysis of the proposed QoS-aware DSA algorithms is also provided.

Introduction

Current wireless networks are characterized by static spectrum allocation and limited user coordination, resulting in very low efficiency in radio spectrum utilization. The emerging paradigm of dynamic spectrum access (DSA) shows promise in alleviating today’s spectrum scarcity problem by ushering in new spectrum-agile networks [1], [2]. Equipped with cognitive radios, users in a network can sense and utilize the available spectrum adaptively [1]. In such an open spectrum approach, each user faces the intricate tradeoff between minimizing interference and maximizing spectrum utilization. This challenging DSA issue is further exacerbated in distributed networks where there is little or no central control over the allocation of wireless resources across users.

The key idea of DSA mechanisms is to dynamically adjust the transmission parameters (e.g., transmission power, channel and rate) of each user according to the behavior of other users sharing the same spectrum bands. Cognitive radios may utilize spectrum based on negotiated or opportunistic spectrum sharing, depending on whether the network is infrastructure-based or ad hoc (any-to-any), and whether the network comprises users of comparable requirements and access opportunities or follows a primary–secondary hierarchy in which primary users hold licenses for the spectrum. Secondary users would access spectrum opportunistically, when they determine that doing so would not adversely affect any primary spectrum holders. This paper deals with DSA among equally-coexisting users in open spectrum systems without a central spectrum controller.

From an information-theoretic viewpoint, the spectrum utilization efficiency of a radio can be measured by the achievable capacity, which is in turn determined by its received signal to interference and noise ratio (SINR) as well as its occupied spectrum bandwidth. In a distributed cognitive radio network, each radio decides on its transmission power and channels based on the sensed radio environment. Its decision not only impacts its own achievable capacity, but also affects that of its neighboring radios via negative interference. Hence, radio resource allocation is an interactive decision-making process, which can be suitably modeled as a multi-player game [2], [17]. Cognitive radios are game players, each of which adopts a capacity-related utility function and takes actions on the transmission power and spectrum occupancy from the action space consisting of available spectrum and allowable power. The tradeoffs between efficient spectrum utilization and interference control can be reflected in the formulation of the utility function employed by each game player, for example, via the introduction of the notion of “interference price” [6].

However, most existing game-based DSA methods aim at enhancing the overall network efficiency, defining the figure of merit to be the total system utility achieved by all users. As such, unbalanced channel allocation is likely to arise, that is, some users gain large portions of the total system utility whereas others get unfairly treated with little spectrum share. This issue is aggravated in a heterogeneous network consisting of users with diverse application-specific QoS requirements. Based on the existing figure of merit, a naive DSA scheme might allocate over-large resources to some users, exceeding the needs for their intended transmissions, while some other users might receive meager capacity below the minimum for successful transmissions. In both cases, the user utility corresponding to its allocated power and spectrum falls outside the acceptable range specified by the user-specific QoS, giving rise to radio resource wastage.

The objective of this paper is to develop distributed DSA solutions that efficiently utilize spectrum with QoS awareness for the distributed networks. We introduce QoS information into the game model to avoid spectrum wastage by proposing two DSA strategies: the QoS-ps-DSA algorithm performs external power scaling to modify the local decision made by each user in order to meet its QoS objective, and the QoSe-DSA algorithm embeds the QoS factor into the utility function so as to use QoS-aware strategies during gaming. The interference price concept is also borrowed to construct a secondary local objective for interference management. Simulations confirm that the proposed DSA techniques offer efficient spectrum sharing in heterogeneous networks with QoS constraints.

The rest of the paper is organized as follows. Section 2 briefly reviews the relevant DSA solutions in the literature. Section 3 describes the system model, while Section 4 gives a brief summary of the QoS-blind DSA algorithms, SC-ADP and MC-ADP [24], [25], [26]. Section 5 presents our proposed QoS-aware DSA algorithms based on distributed games. Section 6 analyzes the convergence properties of the proposed algorithms to Nash Equilibriums and in the limit of the number of users. Section 7 provides corroborating simulation results, followed by concluding remarks in Section 8.