This paper briefly introduced the structure of heterogeneous cellular network and two algorithms which are used for optimizing the network resource allocation scheme: dynamic game algorithm based on spectrum allocation and the game allocation algorithm based on power allocation and alliance. After that, the two algorithms were simulated in MATLAB software and compared with another power iterative allocation algorithm based on non-cooperative game. The results showed that the system energy efficiency of the three algorithms decreased with the increase of the number of small base stations in the network; with the increase of the number of users in the network, the system energy efficiency of the three algorithms increased; under the same number of small base stations or users, the system energy efficiency of the allocation scheme based on the power allocation and alliance was the highest, followed by the power iterative allocation algorithm and the spectrum allocation based game algorithm; the allocation scheme based on power allocation and alliance could provide users with more satisfactory and fair services.
The development of communication technology makes the wireless network carry more and more data, and the development of communication equipment manufacturing industry makes the function of mobile terminal more and more abundant and the cost lower and lower; the former enables wireless network to bear more function applications, and the latter is conducive to the popularization of wireless Internet and mobile terminal [
The principle of cellular network can be simply summarized as user sending information, base station receiving and forwarding information, another base station receiving and sending information, and another user receiving information, which ensures the long-distance transmission of radio waves. Each base station has a certain coverage. When the information of users exceeds the coverage, it will “relay” through another base station. These base stations together constitute the wireless signal transmission network, i.e., cellular network [
Although the cellular network which is composed of base stations [
For the convenience of research and explanation, the double-layer heterogeneous cellular network in
where
I Initialization is performed, and random spectrum is allocated to MBS and SBS;
II The number of users accessing MBS and SBS is collected, and the power and average channel gain of MBS and SBS at the same moment are measured;
III Optimal
IV Step II and III repeat at different moments to get the real-time optimal
The double-layer heterogeneous cellular network in
where
The algorithm adjusts the transmission rate by adjusting the transmission power of SBS in different subchannels, so as to adjust the system energy efficiency of the network. The ultimate goal is to maximize the system energy efficiency. The basic principle of the power allocation and alliance based game algorithm is as follows. Firstly, different SBS in the network forms an alliance. Then, under the alliance structure, the optimal power allocation is solved by non-cooperative game [
I The alliance structure of SBS in the network is initialized. Each SBS is regarded as an alliance.
II After selecting a SBS, n, it is randomly added to an alliance. Then
where
III Then, the optimal power allocation under the alliance structure is solved by non-cooperative game iteration. Firstly, the lower bound power satisfying the user rate is calculated, and the formula is as follows:
where
IV
Power is iterated according to the iterative formula:
where
V Step IV repeats until
VI The system energy efficiency is calculated according to
VII Steps II, III, IV, V and VI repeat for all SBS in the network until SBS is no longer transferred. The power allocation strategy and alliance structure are output after the alliance structure is stable.
In this study, the above two wireless network resource allocation algorithms based on game theory were simulated by MATLAB software [
The simulated heterogeneous cellular network which was used for the two network resource allocation algorithms was similar to
Parameter | Maximum power of MBS | Maximum power of SBS | Noise variance |
Total bandwidth |
---|---|---|---|---|
Numerical value | 80 W | 20 W | –104 dBm | 80 MHz |
Parameter | Number of subchannels | MBS channel gain | SBS channel gain | |
Numerical value | 60 | 1.2 | 1.0 | 0.15 |
Parameter | ||||
Numerical value | 0.4 W | 1.5 Mbps | 1 |
The experimental items which were used for testing the effect of network resource allocation algorithm are as follows.
(1) The number of users participating in the heterogeneous cellular network was set as 50, and the number of SBS in the heterogeneous network was set as 8–15. The system energy efficiency under different number of SBS was tested.
(2) The number of SBS in the heterogeneous network was set as 10, and the number of users participating in the network was set as 20, 30, 40, 50 and 60 respectively. The system energy efficiency under different number of users participating in the network was tested.
(3) The number of SBS in the network was set as 15, and the number of users participating in the network was set as 15, 25, 35, 45 and 55 respectively. Moreover the satisfactory throughput of each user was set, i.e., the satisfactory minimum transmission rate was 2.5 Mbps. The satisfaction of users and the fairness of the network system under different number of users was tested [
where
The above experiments repeated for 5 times under every setting condition, and the average values were taken as the final results. Moreover, in order to further verify the optimization effect of the power allocation and alliance based game algorithm on network resource allocation, in addition to comparing the above two algorithms, they were also compared with an allocation algorithm based on non-cooperative game, i.e., the algorithm described in Section 3.2 which did carry out SBS alliance, but only carried out power iterative optimization.
Different number of SBS were set in the heterogeneous cellular network, and then three algorithms were used for allocating resources. The system energy efficiency after allocation is shown in
Different number of users was set in the heterogeneous cellular network, and the number of SBS was set as 10. Then network resources were allocated using the three algorithms, and the system energy efficiency under the final allocation schemes is shown in
In the heterogeneous network, the number of SBS was set as 15, and different number of users were set. Then network resources were allocated by the three algorithms. The average satisfaction and fairness index of transmission services provided by the heterogeneous network under the final allocation scheme are shown in
This paper briefly introduced the structure of heterogeneous cellular network and two algorithms which were used for optimizing the network resource allocation scheme: the spectrum allocation based dynamic game algorithm and the power allocation and alliance based game allocation algorithm. Then the two algorithms were simulated in MATLAB software and compared with a non-cooperative game based power iterative allocation algorithm. The results are as follows: (1) Under the fixed number of users, the system energy efficiency of the three algorithms decreased with the increase of the number of SBS in the network; under the same number of SBS, the system energy efficiency of the power allocation and alliance based game allocation algorithm was the highest, followed by the non-cooperative game based power allocation algorithm and the spectrum allocation based dynamic game algorithm; (2) Under the fixed number of SBS, the system energy efficiency of the three algorithms increased with the increase of the number of users in the network, but the increasing amplitude decreased gradually, and the satisfaction and fairness decreased with the increase of the number of users; under the same number of users, the system energy efficiency, satisfaction and fairness of the power allocation and alliance based game distribution algorithm were the highest, followed by the non-cooperative game based power allocation algorithm and the spectrum allocation based dynamic game algorithm.