^{1}

^{1}

^{*}

^{2}

In a cognitive radio network wherein primary and secondary users coexist, an efficient power allocation method represents one of the most important key aspects. This paper provides a novel approach based on a game theory framework to solve this problem in a distributed and fair way. Formulated as an optimization problem, the resource allocation problem between secondary users and primary users can be modeled and investigated with the Game Theory, and in particular S-Modular Games, since they provide useful tools for the definition of multi objective distributed algorithms in the context of radio communications. This paper provides also a performance comparison among the proposed game and two other algorithms, frequently used in this context: Simulated Annealing and Water Filling.

Cognitive Radio represents a promising paradigm aimed to optimize the radio spectrum efficiency. In a cognitive radio network, two kind of users can exist: primary (noncognitive) users and secondary (cognitive) users. Even though primary and secondary users coexist within the same network and sharing the same frequency bands, primary users may be unaware of the presence of secondary users. Contrary to primary users, secondary users are smart, since they are intelligent and interact with selfish network users, choosing best operating parameters on the base of the sensed spectrum. Due to the natural radio environment changes, spectrum sharing schemes change frequently, accordingly with the users allocated resource. In this scenario, a game theoretic framework allows to study, model and analyze cognitive radio networks in a distributed way. Such attractive feature allows to achieve the flexibility and the efficient adaptation to the operative environment that were previously mentioned. Due to the players behavior, noncooperative game theory is closely connected to mini/max optimization and typically results in the study of various equilibria, most notably the Nash equilibrium [

Non-cooperative games have been proposed for spectrum sharing in [

This paper extends the above described results providing a distributed game-theoretic approach to obtain a quasi-optimal power allocation method that maximize the energy efficiency of each user, within the coexistence of primary and secondary users. The proposed method take into account throughput fairness among secondary users.

The paper is organized as follows: in Section 2 the proposed system model and applicative scenario are presented. The game description and the NE existence and uniqueness is discussed in Section 3, while in Section 4 the Water-Filling algorithm and a his energy efficient modified version is reported. In 5 the results from computer simulation are commented. Finally some conclusions are expressed in Section 6.

In this work we consider a Cognitive Radio context inspired by a tactical/military scenario where a primary system (owner of the spectrum rights) coexisting with one or more secondary systems and sharing the same frequency band. This kind of situation is very interesting and, at the same time, very frequent, i.e. during coalition deployment of forces or in case of coexistence of humanitarian and military convoys, especially when mobility is taken into account. It is to be noted that such kind of context is a suggested scenario by EDA and NATO [

In the proposed system, each user is characterized by a dedicated sender and receiver, thus each communicating couple consists of a transmitter site and a receiver site, as shown in

the next iteration they can’t do anything else. Each user broadcast a pilot signals at the first iteration of the algorithm in order estimate the channel gain coefficients, that are assumed not changing during the execution of the algorithm.

Since in the above described scenario primary users may be unaware of the presence of secondary users, in the proposed system there can’t be a “direct” cooperation among primary and secondary users. However, by definition, primary users should not undergo a degradation of the required QoS due to the presence of secondary users. For this reason, we propose the following the solution: the primary AP selects and broadcasts periodically on the shared channel a reasonable interference cap on the total interference it willing to tolerate. Together with the interference cap, the value of the total interference received by the primary receiver is transmitted. Thanks to this solution, we introduce a sort of “indirect” unaware cooperation among the two kind of users. The direct result of the introduction of a chosen interference cap is a limitation of the total transmit power of the secondary users on the shared channel. Thanks to the introduction of the interference cap, for the simplicity of exposition, hereinafter we will consider only one primary transmitter-receiver pair, since the proposed scheme can be easily extended to include more than one primary user.

Viewed from the perspective of secondary users, each of them will choose the more suitable transmission power in order to achieve the best transmission quality, respecting the interference cap (broadcasted by primary AP) and ensuring low interference to other secondary users. Due to the consistent decisions made by primary and secondary users, game theory represents an inbred framework to study, analyze and predict the behavior of this system. For simplicity of exposition, we will consider a fixed primary interference cap and therefore a fixed maximum transmission power for the secondary users; this assumption can be made without altering the validity of the system, since variations of this value are relatively slow compared with the time of convergence of the algorithm. In case of wireless networks with high primary mobility and/or more strict delay needs, a delay efficient approach should be followed, see [

The non-cooperative game proposed in this paper can be modeled as game with secondary users, namely the players of the game, operating on one radio resource. This game can be easily extended considering radio resources (i.e. subcarriers of the same multi-carrier channel or different channels) following the approach proposed in [

• Set of Players: where is the -th secondary user.

• Set of Strategies:.

• Utility function: where is the -th secondary user.

Following the approach proposed by [8,13], we take into account the energy efficiency problem at the physical layer, considering an utility function expressed in bit/Joule as performance measure of the model. Each player tries to maximize the following utility function:

where is the complete set of strategies of all secondary users, is the ratio between the number of information bits per packet and the number of bits per packet, is the transmission rate of the th user in bits/sec, is the is the efficiency function (depending on the considered modulation), that represents a stochastic modeling of the number of bits that are successfully received for each unit of energy drained from the battery for the transmission.

Thanks to the efficiency function, the utility function each user tries to maximize is related to its instantaneous signal to noise plus interference ratio (SINR), defined as:

where with the notation we refer to all components of not belonging to user, is the power allocated from the secondary transmitter (),is the path gain from to, is the interference channel from the primary to the secondary receiver, is the primary transmitted power, while is the AWGN component at.

represents the total interference received by the th user and it can be wrote as:

The path gain can be written as [

where and.

The adopted channel model is composed by a small scale fading and a path-loss component. In particular, the path-loss model is the Okomura-Hata model, while the small scale fading is modeled as a Rayleigh process.

Since the above defined utility function depends on the path gains, each secondary user need to know it. In order to solve this problem that could have a strong impact on the signalling process, we assume that each receiver periodically send out a beacon, thanks to which transmitters can measure path gains.

In order to make the NE of the game as efficient as possible (moving it to the Pareto Optimum), we consider the adaptive pricing function that generates pricing values basing on the interference generated by network users. Thus, users that cause high interference transmitting at high power will obtain high value of pricing, due to the fact that is strictly increasing with. The pricing function is written as follows:

where:

• is the maximum pricing value• is the price weight of the generated interference• is the sensitivity of the users to interference.

These three parameters are very useful to adapt the pricing function to the considered wireless network requirements; i.e. we can make the algorithm converge faster decreasing the value of or force all secondary users to transmit at lower power levels increasing their sensitivity to the interference [

A Nash Equilibrium [

Supermodular Games are an interesting class of games that exhibits strategic complementarity. There are several compelling reasons like existence of pure strategy Nash equilibrium, dominance resolvability, identical bounds on joint strategy space etc. that make them a strong candidate for resource allocation modeling. Supermodular games are based on the concept of “supermodularity”, which is used in the social sciences to analyze how one agent’s decision affects the incentives of others.

S-Games are normal form games where is the set of users, the strategy space, the set of utility functions and these conditions are satified:

1) The strategy space of user is a complete lattice.

2) is supermodular in.

3) presents increasing differences in.

The proposed utility function in Equation (1) can be easily demonstrated to be supermodular, since:

1) The strategy space P is a complete lattice;

2)(6)

for all and.

3) The utility function has the increasing difference property.

For the details of proofs we refer to [

Water filling is a frequently used algorithm in power allocation methods. This algorithm starts from the idea that a vase can be filled by a quantity of water equal to the empty volume of the vase. It is well-known in the literature that power allocation in parallel uncoupled channels can follow the water filling principle in order to maximize data-rate. A channel can be filled by an amount of power depending on the existing noise level. A multiuser scenario cannot be modeled as the parallel uncoupled channels case, but it has to be modeled following the approach of an interference channels.

On the base of these considerations, an iterative water filling procedure can be obtained; each user updates its transmission power level as follows:

where is the power level assigned at the user in the iteration and is maximum power that can be transmitted in the channel (the water level). Because of, if, then is assigned to the user.

Iterative Water-Filling gets excellent performances in presence of low interference environments and/or limited number of users. However for increasing values of interference, the algorithm get worst; indeed, users experimenting the best channel conditions will transmit at high power levels, while users experimenting bad channel conditions (i.e. being the receiver close to another transmitter) will receive high interference values and then they will be inactivated. For this reason, EEIWF turns out to be unfair.

For a fixed target data-rate, we can identify a minimum target value for the SINR. In this case, Iterative Water-Filling is energy inefficient, due to the fact that the algorithm tries to maximize the total transmission power, achieving SINR values that are greater than the target value. For this reason, we propose the following energy efficient modified version of the algorithm, called Energy Efficient Iterative Water-Filling (EEIWF). For each iteration, is updated as follows:

• If,

• If,

where represents the reduction factor and it controls the convergence speed of the algorithm. Note that for the algorithm becomes the bisection method.

Such approach allows us to maintain the fixed datarate, using the lowest total transmission power level, taking into account its trend in

In the case of number of user, the SINR trend for decreasing values of is a monotonous decreasing function. Otherwise, when, a reduction of should also improve SINR final value.

In this paragraph we show the results of the simulations that we run in order to verify the behavior of a cognitive network based on the our proposed game-theoretical framework. In the subsection 5.1 the convergence of the algorithm is shown, while in subsection 5.2 a comparison between proposed game and heuristic power allocations will be presented.

The operating context is a terrain square area of 1 km edge, with a suburban path-loss profile. Primary transmitter and receiver positions are fixed; secondary transmitters are independently located in the area, while the secondary receivers positions are placed randomly in a 200 m diameter circle around the respective transmitters. Each secondary user transmits isotropically with , where on the base of a fixed interference cap. Moreover, we consider a noise power, frequency, , a common rate, , and.

Results of a simulation with a primary user and secondary users show a fast convergence of the transmitted power levels and SINR experimented by secondary users. For increasing numbers of secondary users in the networks, the algorithm still maintains a very short time of convergence, see

In order to obtain a qualitative evaluation of the proposed game, we decide to compare its performance with both EEIWF and an optimal centralized heuristic power allocation system, like Simulated Annealing (SA) [

increasing number of secondary users, while all the other parameters of the system remain the same of the previous shown configuration. The simulation results illustrated in

In addition to the SINR, the energy efficiency of the three considered methods is an another important key feature that we need to investigate. If the SINR performance are quite the same for the proposed game and the SA, on the contrary we can observe a great difference in terms of power allocations. Indeed,

In this paper we provide an energy efficient game theoretic framework to solve the resource allocation problem in a cognitive network, wherein primary and secondary users coexist. The power allocation problem is solved thanks to the application of S-Modular Games. Transmission power of secondary users is upper bounded by the interference cap, defined as the total interference

that primary users willing to tolerate, without loosing their required QoS. Moreover, secondary users are discouraged to transmit at high power levels, since they are charged on the base of the interference they generate, thanks to the introduction of a pricing function inside of the utility function. Tuning utility function parameters, the proposed game is able to adapt his performance (in terms of time of convergence) to every kind of network configuration. Indeed, simulation results show a fast convergence of the algorithm for any number of considered users in the cognitive network.

In this work, a performance comparison among the proposed game, an optimal centralized resource allocation method (Simulated Annealing) and an Energy Efficient version of the Water Filling is also included. Simulation results show clearly that game theory obtains better performance than water filling and the proposed game converges to the same SINR values obtained from the heuristic optimization method. However, unlike these, the proposed game results to be the most energy efficient, also for a large number of considered users. Further investigations will be made in order to quantify and analyze the signaling process among secondary users.

The work presented in this paper is part of the CORASMA project (COgnitive RAdio for dynamic Spectrum MAnagement) promoted the European Defense Agency.

We would like to thank Pierpaolo Piunti for the contribution in Water-Filling algorithm.