“Load balancing and Cross-layer resource allocation adapted to C-RAN/VRAN architecture”
Futures cellular wireless networks (fifth generation) are expected to be more heterogeneous with dense deployment of small cells (HetNet), enhanced fronthaul and core architectures and decoupled user traffic and control/management operations . Virtualization and cloud radio access networks (C-RAN) emerges as a promising 5G architectural solutions.
The overall architecture of 5G networks will be more likely composed of a large number of low-cost remote radio heads (RRHs), connected to a base band unit (BBU) pool through high-performance front-haul links. The Common Public Radio Interface (CPRI) and the Open Base Station Architecture Initiative (OBSAI) are the two main specifications for the transport of fronthaul traffic, both are based on digital radio over fiber .
Therefore, traditionally centralized or co-located functionalities have to be split over the C-RAN.
This will lead to many enhancements in terms of network capacity and efficiency:
- First, by moving RRHs closer to the users, a higher system capacity and lower power consumption can be achieved since the signal doesn’t need to propagate a long distance to reach the users.
- Second, migrating the higher layer into the cloud provides an improvement on backhaul/fronthaul latency and throughput (evolution from non-ideal to ideal backhaul), what is desirable for feedback-based algorithms such as Coordinated MultiPoint (CoMP);
- Third, since the baseband processing is centralized at the BBU pool, the cooperative processing techniques to mitigate or to exploit interferences can be leveraged.;
- Forth, by exploiting the resource pooling and statistical multiplexing gain, CRAN is much more efficient in both energy and cost aspects because it is needless to dimension the computing resource of each traditional BS according to the individual peak load.
However, many new challenges rise for resource allocation and scheduling and for the coordination of the distributed overall C-RAN network. Indeed, most of existing optimization approaches and algorithms are no longer adapted to this new architecture.
The above evolutions and properties, added to the fact that C-RAN has different power consumption criteria compared to the conventional networks, provide new motivations and challenges for network optimization issues like cross-layer, coordinated multipoint optimization (CoMP) as well as load balancing in the network.
The aim of the thesis is to conduct performance evaluation studies and to propose enhancements and/or new approaches dealing with the following two topics:
2-1 Load-balancing joint with CoMP in the context of C-RAN
Compared to the fourth generation (4G) cellular systems, the fifth generation wireless communication systems (5G) are anticipated to provide spectral and energy efficiency growth by a factor of at least 10, and the area throughput growth by a factor of at least 25. To achieve these goals, a heterogeneous cloud radio access network (C-RAN) is presented as the advanced wireless access network paradigm, where cloud computing is used to fulfill the centralized large-scale cooperative processing for solving the problem of co-channel interferences.
In addition, the low latency performance guaranteed by C-RAN provides the means to make algorithms such as Coordinated MultiPoint (CoMP) to be implemented. The CoMP algorithms not only provide a solution for the co-channel interference but also introduce new criteria for load balancing  in the network. Having this in mind, in this study, we are searching for CoMP algorithms which improve not only spectral efficiency but also the network stability .
The stability region is defined as the closure of the set of arrival rates at which the network is stable, all users have non-zero throughputs. We are especially interested in the works which account for network dynamics in flow-level , i.e., as the actual set of active users in a network is dynamic and varies as a random process as new data flows are initiated and others complete. A CoMP mechanism that improves the network capacity (Network Stability) is, for example, the one which minimizes the maximum load over all cells involved in a region (HetNet scenario). The objective is to propose CoMP algorithms with performance models and analysis in the context of C-RAN.
2-2 Cross-layer resource allocation to guarantee QoS and spectral efficiency
Decoupling the baseband signal processing from the RRHs is the most attractive feature of C-RAN, which means that RRHs only need to keep the basic transmission and reception functionalities, while computationally intensive tasks can be migrated to the BBU pool in a cloud data center. This centralized signal processing and scheduling feature in the BBU pool further makes a variety of prospective technologies feasible, including centralized encoding and decoding, centralized compression and decompression, and joint beamforming .
Although C-RAN makes it possible to transition conventional cellular networks (CCNs) from hardware defined infrastructures to a software defined environment, many design and operational challenges that have been resolved in CCNs need to be revisited and optimized for C-RAN.
One particular example of importance is the resource allocation problem. Specifically, in CCNs, power control and beamforming strategies have been used to minimize the system power consumption such that users’ predefined quality-of-service (QoS) requirements are fulfilled. Unfortunately, these strategies cannot plug directly into the C-RAN framework. In CCNs, the BSs’ computation capacity is fixed. As a result, resource allocation methods in CCNs are oblivious to the computation capacities of the BSs although users’ achievable QoS levels are actually dependent on them. Under the C-RAN architecture, the computational functionalities in conventional BSs are migrated to the cloud based virtual machines (VMs) in the BBU pool, whose computation capacity can be scaled according to users’ QoS requirements and various parameters from different layers of the OSI stack, including the incoming traffic rate from the application layer and wireless channel state information from the physical layer. Therefore, developing a cross-layer resource allocation scheme is required in order to fully utilize the features of a C-RAN, and to optimize the overall system power consumption. In total three power components can be considered: the power consumption in BBU pool, the power consumption in the fiber links, and the power consumption at the RRHs.
3 REQUIRED SKILLS
The candidate must be graduated from an engineering school and / or with a Master 2 degree whose training focuses primarily on wireless telecommunication systems. Aptitude for teamwork, good spoken and written English will be appreciated.
4 PRACTICAL ISSUES
The PhD student will be hired by Telecom SudParis for the time duration of the PhD (3 years) and registered at the Doctoral School of University Paris-Saclay. Research activities will be conducted in the context of an industrial partnership with Davidson (www.davidson.fr).
5 HOW TO APPLY
Send email to Badii (dot) Jouaber (at) Telecom-SudParis (dot) eu AND to hind (dot) castel (at) telecom-sudparis (dot) eu
- Al-Dulaimi, A. Anpalagan, M. Bennis and A. V. Vasilakos, « 5G Green Communications: C-RAN Provisioning of CoMP and Femtocells for Power Management, » 2015 IEEE International Conference on Ubiquitous Wireless Broadband (ICUWB), Montreal, QC, 2015, pp. 1-5. doi: 10.1109/ICUWB.2015.7324392
- W. Dawson, M. K. Marina and F. J. Garcia, « On the Benefits of RAN Virtualisation in C-RAN Based Mobile Networks, » 2014 Third European Workshop on Software Defined Networks, Budapest, 2014, pp. 103-108. doi: 10.1109/EWSDN.2014.37
- Davydov, G. Morozov, I. Bolotin and A. Papathanassiou, « Evaluation of Joint Transmission CoMP in C-RAN based LTE-A HetNets with large coordination areas, » 2013 IEEE Globecom Workshops (GC Wkshps), Atlanta, GA, 2013, pp. 801-806. doi: 10.1109/GLOCOMW.2013.6825087
- Ishii, Y. Kishiyama and H. Takahashi, « A novel architecture for LTE-B :C-plane/U-plane split and Phantom Cell concept, » 2012 IEEE Globecom Workshops, Anaheim, CA, 2012, pp. 624-630. doi: 10.1109/GLOCOMW.2012.6477646
- N. R. Li, J. Li, W. Li, Y. Xue and H. Wu, « CoMP and interference coordination in heterogeneous network for LTE-Advanced, » 2012 IEEE Globecom Workshops, Anaheim, CA, 2012, pp. 1107-1111. doi: 10.1109/GLOCOMW.2012.6477733
- Sun, Q. Gao, Y. Peng, Y. Wang and L. Song, « Interference management through CoMP in 3GPP LTE-advanced networks, » in IEEE Wireless Communications, vol. 20, no. 1, pp. 59-66, February 2013. doi: 10.1109/MWC.2013.6472200
- Kwak, H. Lee, H. W. Je, J. Hong and S. Choi, « Adaptive and Distributed CoMP Scheduling in LTE-Advanced Systems, » Vehicular Technology Conference (VTC Fall), 2013 IEEE 78th, Las Vegas, NV, 2013, pp. 1-5. doi: 10.1109/VTCFall.2013.6692077
- de la Oliva, J. A. Hernandez, D. Larrabeiti and A. Azcorra, « An overview of the CPRI specification and its application to C-RAN-based LTE scenarios, » in IEEE Communications Magazine, vol. 54, no. 2, pp. 152-159, February 2016. doi: 10.1109/MCOM.2016.7402275
- Zakrzewska, D. López-Pérez, S. Kucera and H. Claussen, « Dual connectivity in LTE HetNets with split control- and user-plane, » 2013 IEEE Globecom Workshops (GC Wkshps), Atlanta, GA, 2013, pp. 391-396. doi: 10.1109/GLOCOMW.2013.6825019
- Baranda, J. Núñez-Martínez and J. Mangues-Bafalluy, « Applying backpressure to balance resource usage in software-defined wireless backhauls, » 2015 IEEE International Conference on Communication Workshop (ICCW), London, 2015, pp. 31-36. doi: 10.1109/ICCW.2015.7247071
- Khan, R. S. Alhumaima and H. S. Al-Raweshidy, « Quality of Service aware dynamic BBU-RRH mapping in Cloud Radio Access Network, » Emerging Technologies (ICET), 2015 International Conference on, Peshawar, 2015, pp. 1-5. doi: 10.1109/ICET.2015.7389166
- Ahlem Khlass, Thomas Bonald, Salah-Eddine Elayoubi:Flow-level performance of intra-site coordination in cellular networks. WiOpt 2013: 216-223
- Ahlem Khlass, Thomas Bonald, Salah-Eddine Elayoubi:Performance evaluation of intra-site coordination schemes in cellular networks. Perform. Eval. 98: 1-18 (2016)