We consider three scenarios according to different traffic intensity and different sizes of data.
In the three scenarios, we show each parameter of scenario setting in Table 11.
Table 11: Traffic intensity and data size
Scenario1
Scenario2
Scenario3
In the three scenarios, we show the simulation results in the figures of the three methods as follows.
19
Figure 2: The sum of channel capacity of three methods at different scenarios when
Figure 3: The percentage of two methods compare to brute-force at three scenarios when
In the Figure2 and Figure 3, we combine sum rate of the three scenarios in a graph and show the sum rate of three methods at three scenarios when the constraint is 1. We can see the result of these two figures, the sum rate of the Total-exchange algorithm is very close to the Brute-force method in
4000
20
the all scenarios, but the sum rate of the Level-exchange algorithm only has the scenario 1 close to the optimal solution. The Level-exchange algorithm depends on the channel capacity of the channels, so the percentage of sum rate of scenario 2 and 3 are lower than the scenario 1.
Figure 4: The sum of channel capacity of three methods at different scenarios when
Figure 5: The percentage of two methods compare to brute-force at three scenarios when 4050
21
In the Figure4 and Figure 5, we combine sum rate of the three scenarios in a graph and show the sum rate of three methods at three scenarios when the constraint is 5. We can see the result of these two figures, the sum rate of the Total-exchange algorithm is also very close to the Brute-force method in the all scenarios, comparing with the Figure2 and Figure3. The results of the Level-exchange algorithm are greater than the results of the Figure2 and Figure3. The
Level-exchange algorithm also has the results which is not stable of the sum rate in the scenarios.
Figure 6: The sum of channel capacity of three methods at different scenarios when 4000
4050 4100 4150 4200 4250 4300 4350 4400 4450
1 2 3
Mbps
Scenario
The sum of channel capacity
Brute-force Total-exchange Level-exchange
22
Figure 7: The percentage of two methods compare to brute-force at three scenarios when
In the Figure6 and Figure 7, we combine sum rate of the three scenarios in a graph and show the sum rate of three methods at three scenarios when the constraint is 10. We can see the result of these two figures, the results of sum rate of the Total-exchange and the Level-exchange are greater than the Figure4 and Figure5. The Total-exchange algorithm is also very close to the Brute-force method in the all scenarios. The Level-exchange algorithm also has the same condition which is like the Figure4 and Figure5.
23
Figure 8: The sum of channel capacity of three methods at different in the Scenario 1
Figure 9: The sum of channel capacity of three methods at different in the Scenario 2 4310
24
Figure 10: The sum of channel capacity of three methods at different in the Scenario 3
In the Figure 8, 9, 10 we can see when the from 1 to 10 the sum rate of three methods are ascend, the reason is when the lower range is greater the sum rate will be better, but it’s not fair to the traffic loads of the operators when the from 1 to 10. The total exchange algorithm a very stable algorithm, because it find all of the exchangeable possibilities when the algorithm is executed in the channel exchange step, and it’s an exponential-time algorithm, but its time complexity is still lower than the brute force method. Then, the level exchange algorithm is sufficiently close to the sum rate of the brute force method in the three scenarios. Comparing with the total exchange algorithm, the sum rate of the level exchange algorithm is unstable. Because the level exchange algorithm is sensitive to the channels of channel capacity, but its time complexity isn’t an exponential-time algorithm. It is much faster than the total exchange algorithm, due to when we do channel exchange, the level exchange algorithm will exchange the channels level by level that will reduce the exchange effort.
25
Chapter6. Conclusion and Future Work
In this thesis, we proposed two channel assignment algorithms to solve our problem and reduce the time complexity of the problem which is an NP-hard problem. One is called total exchange algorithm which is also an exponential-time algorithm - comparing with the brute force method- where m is the number of operators and n denoted the number of channels, the time complexity of total exchange algorithm is less than the brute force method, and the throughput of this algorithm is very close the optimal solution of our problem. The other one is called level exchange algorithm which is not an exponential-time algorithm – ([ ] ) comparing with the brute force method and the total exchange algorithm, the time complexity significantly reduce than the other methods, so this algorithm is much faster than the others, and its throughput is also close the optimal solution of our problem.
In the simulations, we compare throughputs of the algorithm of we proposed, and consider three scenarios. The result of simulations has shown that the throughput of our proposed methods not only reduce the time complexity but also close the optimal solution of our problem.
Our algorithms can also extend to the other problem, such as the product scheduling in the Manufacturing Industry, or resource allocation problems … etc. In the future works, we can improve the throughputs of the algorithm and do the mathematical analysis for the methods.
26
REFERENCES
[1]FCC, ET Docket No 03-222 Notice of Proposed Rule Making and Order,Dec. 2003.
[2]J. Mitola, “Cognitive radio: An integrated agent architecture for software defined radio,” Ph.D.
dissertation, KTH Royal Inst. of Technol.,Stockholm, Sweden, 2000.
[3]I. F. Akyildiz, W.-Y. Lee, M. C. Vuran, and S. Mohanty, “Next generation/dynamic spectrum access/cognitive radio wireless networks: A survey,” Comput. Netw., vol. 50, no. 13, pp. 2127–2159, Sep. 2006.
[4] Z. Ji and K. J. R. Liu, “Dynamic spectrum sharing: a game theoretical overview," IEEE Commun.
Mag., vol. 45, no. 5, pp. 88-94, May 2007.
[5] X. Wang, Z. Li, P. Xu, Y. Xu, X. Gao, and H. Chen, “Spectrum Sharing in Cognitive Radio Networks – An Auction-based Approach,” to appear in IEEE Trans. System, Man Cybern, B, Cybern., 2010.
[6] Y. Hou, Y. Shi, and H. Sherali, “Spectrum sharing for multi-hop networking with cognitive radios,” IEEE Journal on Selected Areas in Communications, vol. 26, no. 1, pp. 146–155, 2008.
[7 ]M. Ma and D. Tsang, “Impact of Channel Heterogeneity on Spectrum Sharing in Cognitive Radio Networks,” in IEEE International Conference on Communications, 2008. ICC’08, 2008, pp.
2377–2382.
[8] T. Lv, T. Wang, and X. Yu, “Primary User Activity Based Channel Allocation in Cognitive Radio Networks,” in IEEE Vehicular Technology Conference, 2010, pp. 1–5.
[9] W. Wang and T. Lv, “A Novel Spectrum Sharing Algorithm Based on the Throughput in Cognitive Radio Networks,” IEEE International Conference on Wireless Communications, Networking and Mobile Computing, Wicom 2009.
[10] Tan and L.T. “Channel Assignment for Throughput Maximization in Cognitive Radio Networks,” in IEEE Wireless Communications and Networking Conference, April 2012, pp. 1427 - 1431
[11] J. Xiang, Y. Zhang, T. Skeie, and J. He, “Qos aware admission and power control for cognitive radio cellular networks,” Wireless Commun. and Mobile Comput., 2009.
[12] A. T. Hoang and Y.-C. Liang, “Downlink channel assignment and power control for cognitive networks,” IEEE Trans. Wireless Commun., vol. 7, no. 8, pp. 3106–3117, Aug. 2008.
[13] M. Ma and Tsang D. H. K., “Impact of channel heterogeneity on spectrum sharing in cognitive radio networks,” Communications, (ICC'08) IEEE Intl. Conf. on, pp. 2377-2382, May 2008.
27
[14] J. Tang, G. Xue, C. Chandler, and W. Zhang, “Link scheduling with power control for throughput enhancement in multihop wireless networks,” IEEE Trans. Veh. Technol., vol. 55, no. 3, pp. 733–742, May 2006.
[15] Y.T. Hou, Y. Shi, and H.D. Sherali. Optimal spectrum sharing for multihop software defined radio networks. Proc. IEEE INFOCOM, May 2007.
[16] S. M. Almalfouh and G. L. Stüber, “Uplink resource allocation in cognitive radio networks with imperfect spectrum sensing,” in Proc. IEEE Veh. Technol. Conf., Sep. 2010, pp. 1–6.
[17] Silvano Martello and Paolo Toth, Knapsack Problems: Algorithms and Computer Implementations. John Wiley & Sons, 1990.
[18] Donald Gross, John F. Shortle, James M. Thompson, Carl M. Harris, Fundamentals of Queueing Theory. John Wiley & Sons, 2011.