Multiuser Resource Allocation Algorithms for Downlink OFDMA-based MIMO Network
Abstract
The problem of simultaneous multiuser resource (subcarriers-and-bits) allocation algorithm in OFDMA-based multiple input-multiple output (MIMO) system has recently attracted significant interest. In this paper, we employ adaptive modulation technique and advanced use of multiple antennas at both the transmitter and receiver to develop four resource allocation schemes. The first scheme assigns subcarrier to the user with best channel gain and employs spatial multiplexing (SM) on the MIMO system to further enhance the throughput. The space-division multiple-access (SDMA) scheme assigns single subcarrier simultaneously to the terminals with pairwise “nearly orthogonal” spatial signatures. In the third scheme, we propose to design the transmit beamformers based on the zero-forcing (ZF) criterion such that the multiuser interference (MUI) is completely removed. Specifically, we propose a low-complexity iterated terminals-selection algorithm in conjunction with the ZF criterion such that the selected ZF (SZF) scheme can be exploited to achieve throughput multiplication. Alternatively, we propose a least-squares (LS) based multiuser resource allocation algorithm to cope with the over-determined system such that all users are allowed to share single subcarrier.
Keywords
Download Options
Introduction
In 4G and future cellular networks, mushrooming users need to share the spectrum to achieve high-rate multimedia communication while ensure the fulfillment of quality-of-service (QoS) requirements. In the downlink (DL) of 4G LTE, orthogonal frequency division multiple access (OFDMA) technique is employed [1]. The advantages of OFDMA includes robust multipath suppression, ability to combat intersymbol interference (ISI), relative low complexity, and flexibility in accommodating many users with widely varying data rates [2]. The rationale for OFDMA is dynamically allocating subcarriers to the user with best channel state. However, the resource allocation algorithms are not specified in the LTE standard, and recently, quite a few scheduling and resource allocation algorithms have been proposed for OFDMA cellular system [3-7]. Resource allocation is essentially a constrained-optimization problem [Chap. 18 of [8]] that either maximizes the overall data rate or minimizes the total transmit power subject to specific constraints, e.g., the users’ QoS requirements. The optimum resource allocation algorithm is itself a NP-complete problem whose solution can only be found with exhaustive search [7], which is infeasible for practical situation. In the work of [1], the sum rate of all users is maximized with the constraint of total transmit power and minimum data rate for each user. While the work of [3] aims to allocate the subcarriers and power such that the minimum user’s data rate is maximized. Alternatively, proportional fairness scheduling [4] is designed to take advantage of multiuser diversity, while maintaining comparable long-term throughput for all users. In the work of [5], a Lagrangian-based algorithm is proposed to attain a dramatic gain in power efficiency. However, the load of computation is too high to put it into practical use.It is well-known that one of the largest advantages of LTE over incumbent standard is more advanced use of multiple antennas. And the massive multiple input-multiple output (MIMO) technique is apt to be included in future 5G network [9-12].
Conclusion
In this paper, we have proposed four resource allocation algorithms in OFDMA based multiuser MIMO system, namely, SM, SDMA, SZF and LS schemes. All the proposed algorithms are channel aware in the sense that they adapt to the channel conditions. Specifically, the SDMA and SZF algorithms judiciously select co-channel users based on their spatial signatures to enable significant improvement in the system throughput. Analytical and simulation results demonstrated that employing the antenna array at the BTS and RT has extensively increased the spectral efficiency. Furthermore, since the proposed algorithms are simple yet reliable, it is plausible to apply it for resource allocation in practical multiuser OFDMA-based LTE system, where multiple antenna transmission and reception are available.