A Cross-layer Resource Allocation and Scheduling
for Multiuser Space-Time Block Coded
MIMO/OFDM Systems
Weilan Huang, Student Member, IEEE and K. B. Letaief, Fellow, IEEE
Center for Wireless Information Technology
Electrical and Electronic Engineering Department
The Hong Kong University of Science and Technology
Clear Water Bay, Kowloon, Hong Kong
Email: wlhuang@ust.hk
, eekhaled@ee.ust.hk
Abstract—In this paper, we propose a cross-layer adaptive
resource allocation and scheduling approach for multiuser
MIMO/OFDM systems. Specifically, a space-time block coded
multi-antenna combined with OFDM system is considered where
the channel state information (CSI) is not perfectly known. A
distinctive feature of the cross-layer optimization methodology is
the joint design of the subcarrier, bit, and power allocation in the
physical layer along with the scheduling in the data link layer.
Simulation results will demonstrate that the proposed cross-layer
optimization and design approach achieves significant gains in
throughput, average packet delay, and system stability compared
to a conventional MIMO/OFDM system where the resource
allocation and scheduling are performed without taken into
consideration the cross-layer interactions between the different
protocol layers.
Keywords-MIMO OFDM; cross-layer design; beamforming;
adaptive resource allocation; packet scheduling; partial CSI;
multiple-access
I. INTRODUCTION
Current wireless networks are designed in layers according
to the OSI (Open Systems Interconnection) reference model
with each layer having its own design. Each layer optimizes its
own goal and the design can hardly be optimal from a system
point of view. Because of the increasing demand of wireless
communication systems and limited resources, it is more and
more necessary for the system designer to implement more
efficient protocols through a cross-layer approach [1]. The
nature of cross-layer design is to provide an innovative insight
into the vertical integration of different protocol layers with the
ultimate goal of achieving efficient management of system
resources. In particular, how to utilize smart scheduling
mechanisms to improve efficiency based on instantaneous user
demand and channel state for each user is essential for the next-
generation wireless communication networks.
Orthogonal frequency division multiplexing (OFDM) has
been identified as a very promising technique to efficiently
utilize the limited RF bandwidth and transmit power in
wideband transmission over time-dispersive multipath
channels. Multi-Input Multi-Output (MIMO) technology has
also been recognized as a key approach for achieving a
dramatic increase in the capacity of wireless communication
systems [2]-[3]. In particular, the use of MIMO technology
combined with OFDM is an attractive solution for future
broadband wireless systems. A key advantage of
MIMO/OFDM systems is their inherent structure that allows
the use of dynamic resource allocation where the subcarrier,
bit, and power, for example, can be adapted dynamically [4]-
[6].
A critical element in all of the allocation schemes is the
channel state information (CSI) knowledge. Most of literatures
in the context of adaptive resource allocation assume perfect
CSI at the transmitter [4]-[5]. However, perfect CSI is rarely
available. Many papers have considered the problem of how to
use partial CSI to optimize the transmitter design for single
carrier systems [7]-[8]. Recently, some papers extended this
work to multi-carrier systems [6]. But they mainly dealt with
the single user scenario.
In this paper, we propose a cross-layer adaptive resource
allocation algorithm for MIMO/OFDM systems by utilizing
partial CSI. Based on a transmission scheme that combines the
benefits of transmit beamforming and orthogonal space-time
block coding [6], an adaptive multiuser subcarrier, bit, and
power allocation algorithm is proposed. The allocation results
that encapsulate the channel conditions and power limitation
observed in the physical layer are then passed to the data link
layer. The scheduler in the data link layer controls the order
and number of packets, which takes the allocation results,
packet arrival, quality of service (QoS) requirements, user
fairness and complexity into consideration. According to the
scheduling results, the physical layer then reallocates the
subcarriers, bits, and power dynamically, for the sake of
transmitting these scheduled packets and providing guaranteed
performance simultaneously. This is done so as to 1) to
maximize the throughput for a given power; 2) to ensure
fairness; and 3) to guarantee the QoS requirements to all the
users.
The remainder of this paper is organized as follows. In
Section II, the system model is described. The unconstrained
allocation problem and packet-scheduling algorithm are
presented in Section III. The updated optimization problem and
the cross-layer adaptive allocation scheme are studied in
Section IV. In Section V, we investigate the performance of
the proposed system. Finally, we conclude the paper in Section
VI.
This work is supported by the Research Grant Council Under Grant #
HKUST6250/04E.
0-7803-8939-5/05/$20.00 (C) 2005 IEEE