School of Electrical and Computer Engineering Georgia Institute of Technology
Broadband Wireless Networking Lab
5G Cellular Systems
Project Description
The evolution from 4G to 5G systems will make possible a number of deployment scenarios that haven't existed in 4G systems. These deployment scenarios will play a great role in defining the technologies that will drive 5G innovation.
5G cellular systems are expected to utilize new frequency bands including the 5GHz unlicensed bands through the use of intersite carrier aggregation or dual connectivity. More significantly, the use of ultra-high frequencies including mmwave and terahertz to realize peak data rates of over 10Gbps is also being investigated rigorously. These mmwave or terahertz small cells are also expected to support a massive multiple-input multiple-output (MIMO) antennas in order to overcome the high pathloss and blocking associated with such high frequencies.
5G systems are also expected to leverage the different radio access technologies and provide a unified architecture to exploit transmission opportunities across several RATs in a robust and seamless fashion.
Another striking scenarios will be the deployment of massive number of machine type devices. These devices can further be classified into low-latency, low-power and mission critical devices. Such a wide-range of devices will require the use of a multitude of technologies that will be part of 5G systems.Ad hoc deployments such as device-to-device communications and inter-vehicular and vehicle-to road communications are also envisioned as part of the 5G deployment scenarios. The different deployment scenarios for 5G systems are illustrated below.
|
| Deployment Scenarios envisioned for 5G Cellular Systems |
This project will investigate the key technologies envisioned to play a major role in shaping the 5G cellular systems. This includes a thorough analysis of massive MIMO systems, 3D channel modeling, mmWave and Terahertz communications, multi-layer multi-RAT HetNets and energy efficient communication. These are discussed in detail below.
The MIMO techniques are expected to pave the way for supporting the ever increasing demands of higher date rate transmissions in 5G systems. In the mean time, several candidate MIMO architectures arise for 5G cellular networks, including massive MIMO, mmWave MIMO and network MIMO, which promise significant gains with higher reliability.
Massive MIMO enabled by a large number of antennas at the base stations is termed as an energy-efficient way to serve a large number of users simultaneously and improve the system capacity. Both centralized and distributed architectures will be considered. However, the challenges arising from channel estimation, pilot contamination and inter-cell interference are still open. To this end, novel hybrid TDD-FDD architecture will be investigated to minimize the inter-cell interference as well as the pilot contamination for massive MIMO systems.
While the critical shortage is getting more and more serious, huge amount of spectrum is available in mmWave bands. Besides, the MIMO architecture in mmWave bands is able to offset the high path loss via its high directional transmission. Another benefit is the lower interference in mmWave MIMO due to the narrow beamwidth. On the other hand, complete understanding of the EM wave propagation characteristics in mmWave band is required for mmWave MIMO channel modeling. Other challenges, such as the hybrid beamforming methods, also should be investigated
In conventional multiuser MIMO system, the inter-cell interference limits the full utilization of the spatial multiplexing gain. Hence, the cooperated MUMIMO, also known as network MIMO, is deemed as a key candidate to exploit the intercell interference and improve the system throughput. Specifically, the network MU-MIMO is achieved by BS coordination or clustering, where the transmitting antennas of a cluster of BSs cooperatively act as a single antenna array and each user in the cluster may receive useful signals, instead of interference, from several nearby BSs. Despite of its significant advantages, the fundamental challenges should be addressed for the realization of Network MIMO, such as cluster edge effects, backhaul constraints, system overhead.
Increasing the number of cells over a given area has the benefits of increasing the network capacity and the area spectral efficiency by several orders of magnitude. 4G systems have already witnessed the rise of a number of small cells overlaid on the existing macrocell area. Although small cells promise significant network capacity gains, they have led to several challenges in terms of interference management, mobility management and energy consumption.
The spatial reuse of frequency introduces the fundamental problem of inter-cell interference (ICI) in the network caused by the sharing of the spectrum among the different tiers of the HetNets. Our initial work involved developing the analysis and mitigation techniques for the fundamental problem of interference in multi-layer HetNets. Spatial statistics tools were utilized that allow the reconstruction of complete coverage maps. To this end, a novel correlation analysis has been conducted by deriving a spatial coverage cross-tier correlation function. Based on this, correlation-aware cell biasing has been conducted to mitigate inter-cell interference.
Secondly, frequent handovers become a major issue in multi-layer HetNets. The key goals of handover management for HetNets are to minimize the handover failures, number of handovers and the signaling overheads associated with it. A local anchor based scheme is a promising approach where a chosen small cell can act as a local anchor for a cluster of small cells. Intelligent traffic-aware admission control algorithms also benefit in determining the optimal traffic offload from the macrocell to the small cells.
In addition to the multi-layer case, 4G systems have focused on the interworking between the different RATs. However, a unified architecture to exploit all the RATs to achieve maximum gains is only becoming a part of 5G systems. Several new architectures are currently being investigated for multi-layer multi-RAT HetNets. Multi-stream carrier aggregation and dual connectivity where the macrocell acts as anchor cell providing coverage and small cells act as capacity boost are promising architectures to achieve seamless mobility performance for users
Multi-stream carrier aggregation, however, raises several questions such as handling an explosive number of frequency measurements in order to discover the small cells. Furthermore, determining the best set of small cells or component carriers for the users on the go, based on the traffic and mobility statistics are very crucial to achieve superior mobility performance.
Futhermore, the adoption of new RAT in 5G systems, such as the mmWave air interface, also leads to several new challenges. Due to the unique propagation characteristics of mmWave cells, a new multi-RAT architecture is required to overcome the high path loss and frequent handovers. One such approach is the concept of phantom cell with a control and data split between the legacy macrocell and mmWave small cell. In spite of the new architectures, a detailed study on the problem of handovers in mmWave cells employing directional beamforming is fundamental to enable such diverse multi-layer multi-RAT systems.
Energy Efficient Communication
The energy consumed in base stations not only represents significant amount of operational expenses, but it is also mostly wasted due to a low energy efficiency.
The traditional approach to addressing such inefficiency has been to improve the hardware components. Alternatively, better network planning and deployment strategies, adapting base stations' active periods according to the traffic demands, and improving the energy communication techniques have shown promising results in achieving energy consumption reduction.
Firstly, the key objective is to model and analyze the energy consumption in 5G cellular systems and develop techniques to minimize it. This will encompass the characterization of all the energy consumed at the base stations. The second objective will be to utilize a novel on-off and cell-association scheme to minimize the overall network energy consumption while satisfying the spatially- and temporally-varying traffic demands.
With the emergence of multi-stream carrier aggregation architectures, novel techniques must be developed not only to improve the energy efficiency, but also to balance it with the conflicting objective of capacity maximization.
Finally, energy efficient transmission and reception at the user equipment will be one of the primary goals of 5G systems. With this in mind, the performance of discontinuous reception methods for energy savings within the user equipments need a thorough investigation. Given that the new scenarios will increasingly support different flavors of carrier aggregation, a cross-carrier-aware technique to further enhance energy savings at the user equipment should also be investigated.
3D End-to-end Channel Modeling for mmWave and THz bands
In order to make full use of full-dimension MIMO, we need to learn how the wave is propagating in both elevation and azimuth plan. Thus, a realistic 3D channel modeling is required. Currently, a reliable and complete 3D channel model is not available yet. In Terahertz band, the molecular absorption is high so we need to employ a highly directional antenna to make sure the received power is at a certain level. Therefore, a 3D end-to-end channel model, which captures the characteristics of wave propagation and antenna directivity needs to be studied in great detail.
You are visitor:
since 11/26/2014.