In this paper, we propose mapped twoway water filling (MTWF) scheme to maximize energy efficiency (EE) for hybrid bursty services with quality of services (QoS) requirements in twoway multirelay (TWMR) OFDM networks. The bursty traffic is first analyzed by strictly proved equivalent homogeneous Poisson process, based on which the QoS requirements are converted into sumrate constraints. The formulated nonconvex EE maximization problem, including subcarrier assignment, relay selection (RS) and rate allocation, is NPhard involving combinatorial optimization. To conduct optimal RS on each subcarrier without priori bursty traffic knowledge, we utilize some approximate relationships under high data rate demands to remove its dependence on twoway data rates, and simplify the whole optimization problem as well. After the optimal channel configuration is obtained, which only depends on channel conditions, subcarrier assignment is attained through elitist selection genetic algorithm (ESGA), and rate allocation of each service is fulfilled by deducing twoway water filling principle. A new equivalent optimization objective function is proposed next as the simple evaluating index in ESGA to reduce complexity. Finally, simulations are carried out to verify the superiority and convergence of our scheme, as well as the applicability for different scenarios.
The work was supported by National Nature Science Foundation of China Project (Grant No. 61471058), Hong Kong, Macao and Taiwan Science and Technology Cooperation Projects (Grant Nos. 2014DFT10320, 2016YFE0122900), Beijing Nova Program (Grant No. xx2012037), Shenzhen Science and Technology Project (Grant No. 20150082), and Beijing Training Project for the Leading Talents in S&T (Grant No. Z141101001514026).
Appendix
Proof of the average performance equivalent
With the same data packet departure (that is homogeneous Poisson process), there are two kinds of data packet arrival, which are equivalent about the average performance, i.e., delay in this paper. The two types of data packet arrival can be described as follows.
Homogeneous Poisson arrival: The data packets arrive with a fixed rate $\lambda$ with a period of time $T$.
Heterogeneous Poisson arrival: The data packets arrive with a changed rate $\lambda(t)$ with regard to time $t$ within a period of time $T$, and the following constraint is satisfied, as given by
Since the departure process is identical, the equivalence of the two queue systems can be ensured if the equivalence of the two arrival processes can be proved within time duration $T$. Observing the whole queue system at time $T$, it is seen that the waiting time, depending on the state of queue system, will be same for different queue systems as long as the probability of queue length is identical. It is noted that only same average performance at time $T$ is guaranteed, but not the transient performance or any time before $T$.
As for homogeneous poisson arrival, the probability of arriving $k$ data packets within time $T$ can be expressed as
Similarly, for heterogeneous poisson arrival, this probability of arriving $k$ data packets can be obtained, as given by
The average time duration can be written as $T_1^s+\frac{1}{\Lambda_1^s}$ for service $s$ in downlink from a burst arrival to another burst arrival periodically. The burst data packet arrival is a heterogeneous poisson process, thus the following relationship can be obtained with the arrival rate $\lambda_1^s$ during bursty duration while $0$ during bursty interval, as given by
Since ${\lambda_1^s}^*=\frac{\lambda_1^s~\cdot~T_1^s}{{T_1^s+\frac{1}{\Lambda_1^s}}}$, the new homogeneous poisson arrival with the parameter of ${\lambda_1^s}^*$ can meet the constraint
Therefore, it is completely proved that it is equivalent on average performance for the bursty arrival and the homogeneous poisson arrival with the parameter of ${\lambda_1^s}^*$.
[1] Daquan Feng , Chenzi Jiang , Gubong Lim . A survey of energyefficient wireless communication. IEEE Commun Surv Tutorials, 2013, 15: 167178 CrossRef Google Scholar
[2] Cui Q, Wang H, Hu P. Evolution of LimitedFeedback CoMP Systems from 4G to 5G: CoMP Features and LimitedFeedback Approaches. IEEE Veh Technol Mag, 2014, 9: 94103 CrossRef Google Scholar
[3] Miao G, Himayat N, Li Y G. Crosslayer optimization for energyefficient wireless communications: a survey. Wirel Commun Mob Comput, 2009, 9: 529542 CrossRef Google Scholar
[4] Liu Y, Lu L, Li G Y. Joint User Association and Spectrum Allocation for Small Cell Networks With Wireless Backhauls. IEEE Wireless Commun Lett, 2016, 5: 496499 CrossRef Google Scholar
[5] Min Zhou , Qimei Cui , Jantti R. Energyefficient relay selection and power allocation for twoway relay channel with analog network coding. IEEE Commun Lett, 2012, 16: 816819 CrossRef Google Scholar
[6] Li Y, Louie R H Y, Vucetic B. Relay selection with network coding in twoway relay channels. IEEE Trans Veh Technol, 2010, 59: 44894499 CrossRef Google Scholar
[7] Peters S, Panah A, Truong K, et al. Relay architectures for 3GPP LTEadvanced. EURASIP J Wirel Commun Netw, 2009, 2009: 114. Google Scholar
[8] Rankov B, Wittneben A. Spectral efficient protocols for halfduplex fading relay channels. IEEE J Select Areas Commun, 2007, 25: 379389 CrossRef Google Scholar
[9] Fu S, Lu K, Zhang T, et al. Cooperative wireless networks based on physical layer network coding. IEEE Wirel Commun, 2010, 17: 8695. Google Scholar
[10] Popovski P, Yomo H. Physical network coding in twoway wireless relay channels. In: Proceedings of IEEE International Conference on Communications (ICC), Scotland, 2007. 707712. Google Scholar
[11] Kim S J, Mitran P, Tarokh V. Performance bounds for bidirectional coded cooperation protocols. IEEE Trans Inform Theor, 2008, 54: 52355241 CrossRef Google Scholar
[12] Cui Q, Yuan T, Tao X. Energy Efficiency Analysis of TwoWay DF Relay System With NonIdeal Power Amplifiers. IEEE Commun Lett, 2014, 18: 12541257 CrossRef Google Scholar
[13] Ngo H Q, Quek T Q S, Shin H. Amplifyandforward twoway relay networks: error exponents and resource allocation. IEEE Trans Commun, 2010, 58: 26532666 CrossRef Google Scholar
[14] Cui Q, Yuan T, Ni W. Energyefficient twoway relaying under nonideal power amplifiers. IEEE Trans Veh Tech, 2016, 99: 12571270. Google Scholar
[15] Kim S, Lee Y H, Shahbazpanahi S. Energyefficient power allocation for OFDM signaling over a twoway AF relay. IEEE Trans Veh Tech, 2014, 64: 48564863. Google Scholar
[16] Sun C, Cen Y, Yang C. Energy efficient OFDM relay systems. IEEE Trans Commun, 2013, 61: 17971809 CrossRef Google Scholar
[17] Pang L, Zhang Y, Gong F. Energy Aware Resource Allocation for Incremental AFOFDM Relaying. IEEE Commun Lett, 2015, 19: 17661769 CrossRef Google Scholar
[18] TainSao Chang , KaiTen Feng , JiaShi Lin . Green Resource Allocation Schemes for RelayEnhanced MIMOOFDM Networks. IEEE Trans Veh Technol, 2013, 62: 45394554 CrossRef Google Scholar
[19] Do T P, Wang J S, Song I. Joint relay selection and power allocation for twoway relaying with physical layer network coding. IEEE Commun Lett, 2013, 17: 301304 CrossRef Google Scholar
[20] Talwar S, Jing Y, Shahbazpanahi S. Joint Relay Selection and Power Allocation for TwoWay Relay Networks. IEEE Signal Process Lett, 2011, 18: 9194 CrossRef ADS Google Scholar
[21] Wang X M, Zheng F C, Zhu P C, et al. Energyefficient resource allocation for OFDMA relay systems with imperfect CSIT. Sci China Inf Sci, 2015, 58: 082311. Google Scholar
[22] Chen Y, Fang X, Huang B. EnergyEfficient Relay Selection and Resource Allocation in Nonregenerative Relay OFDMA Systems. IEEE Trans Veh Technol, 2014, 63: 36893699 CrossRef Google Scholar
[23] Cheung K T K, Shaoshi Yang K T K, Hanzo L. Achieving Maximum EnergyEfficiency in MultiRelay OFDMA Cellular Networks: A Fractional Programming Approach. IEEE Trans Commun, 2013, 61: 27462757 CrossRef Google Scholar
[24] Xiong K, Fan P, Lu Y. Energy Efficiency With Proportional Rate Fairness in Multirelay OFDM Networks. IEEE J Select Areas Commun, 2016, 34: 14311447 CrossRef Google Scholar
[25] Cheung K T K, Yang S, Hanzo L. Spectral and Energy Spectral Efficiency Optimization of Joint Transmit and Receive Beamforming Based MultiRelay MIMOOFDMA Cellular Networks. IEEE Trans Wireless Commun, 2014, 13: 61476165 CrossRef Google Scholar
[26] Zhou M, Cui Q, Valkama M, et al. Energyefficient resourse allocation for OFDMAbased twoway relay channel with physicallayer network coding. EURASIP J Wirel Commun Netw, 2012, 66: 111. Google Scholar
[27] Liew S C. Performance of various inputbuffered and outputbuffered ATM switch design principles under bursty traffic: simulation study. IEEE Trans Commun, 1994, 42: 13711379 CrossRef Google Scholar
[28] Jacob L, Kumar A. Delay performance of some scheduling strategies in an input queuing ATM switch with multiclass bursty traffic. IEEE/ACM Trans Networking, 1996, 4: 258271 CrossRef Google Scholar
[29] Yu M, Zhou M. A Performance Modeling Scheme for Multistage Switch Networks With PhaseType and Bursty Traffic. IEEE/ACM Trans Networking, 2010, 18: 10911104 CrossRef Google Scholar
[30] Jagannathan K, Jiang L, Naik P L, et al. Scheduling strategies to mitigate the impact of bursty traffic in wireless networks. In: Proceedings of International Symposium and Workshops on Modeling and Optimization in Mobile, Ad Hoc and Wireless Networks (WiOpt), Japan, 2013. 468475. Google Scholar
[31] Wu J, Bao Y, Miao G, et al. Base station sleeping and power control for bursty traffic in cellular networks. In: Proceedings of IEEE International Conference on Communications Workshops (ICC), Sydney, 2014. 837841. Google Scholar
[32] Wu Y, Min G, Yang L T. Performance analysis of hybrid wireless networks under bursty and correlated traffic. IEEE Trans Veh Technol, 2013, 62: 449454 CrossRef Google Scholar
[33] Ng T, Doan Hoang T. Joint optimization of capacity and flow assignment in a packetswitched communications network. IEEE Trans Commun, 1987, 35: 202209 CrossRef Google Scholar
[34] Cidon I, Khamisy A, Sidi M. Analysis of packet loss processes in highspeed networks. IEEE Trans Inform Theor, 1993, 39: 98108 CrossRef Google Scholar
[35] Gurewitz O, Sidi M, Cidon S. The ballot theorem strikes again: packet loss process distribution. IEEE Trans Inform Theor, 2000, 46: 25882595 CrossRef Google Scholar
[36] Jianming Liu , Xiaohong Jiang , Horiguchi S. Recursive formula for the moments of queue length in the M/M/1 queue. IEEE Commun Lett, 2008, 12: 690692 CrossRef Google Scholar
[37] Medhi J. Stochastic Models in Queueing Theory. San Diego: Academic Press, 2003. Google Scholar
[38] Miao G, Himayat N, Li G. Energyefficient link adaptation in frequencyselective channels. IEEE Trans Commun, 2010, 58: 545554 CrossRef Google Scholar
[39] Holland J H. Adaptation in Natural and Artificial Systems. Ann Arbor: University of Michigan Press, 1975. Google Scholar
[40] Goldberg D E. Genetic Algorithms in Search, Optimization, and Machine Learning. New York: AddsinWesley Publishing Company, 1989. Google Scholar
[41] Jiang H, Zheng L, Liu Y, et al. Multiconstratined QoS routing optimization of wireless mesh network based on hybrid genetic algorithm. In: Proceedings of International Conference on Intelligent Computing and Integrated Systems (ICISS), Guilin, 2010. 862865. Google Scholar
[42] Hu X M, Zhang J, Yu Y. Hybrid genetic algorithm using a forward encoding scheme for lifetime maximization of wireless sensor networks. IEEE Trans Evol Computat, 2010, 14: 766781 CrossRef Google Scholar
[43] Rudolph G. Convergence analysis of canonical genetic algorithms.. IEEE Trans Neural Netw, 1994, 5: 96101 CrossRef PubMed Google Scholar
Figure 1
(Color online) Hybrid bursty services transmission in TWMR OFDM networks.
Figure 2
(Color online) Bursty traffic of single service.
Obtain sumrate constraints from the QoS requirements through equivalent queue analysis for all services; 
According to the integratednoisechannel selection criterion, the optimal relay for each subcarrier is selected and the optimal channel conditions are recorded; 
Calculate the comprehensive twoway channel coefficients for each subcarrier; 
Initialize the subcarrier assignment sequence satisfying the $\{C_s^n\}$ constraints of each service for chromosome $i$; 


Resolve water level according to 
Conduct the rate allocation and obtain the data rate ${r_1}$ and ${r_2}$ according to water level for service $j$; 

Let these negative data rates be 0 and exclude these subcarriers. Turn to Step 


Calculate the fitness according to the equivalent optimization function ( 

Select and record the best assignment scheme; 
Obtain the next generation by crossover and mutation; 
Copyright 2020 CHINA SCIENCE PUBLISHING & MEDIA LTD. 中国科技出版传媒股份有限公司 版权所有
京ICP备14028887号23 京公网安备11010102003388号