Demonstration of Highly Programmable Downlink OFDMA (WiMax)Transceivers for SDR Systems Hamid Eslami Gaurav Patel Chitaranjan P.Sukumar Sang V. Tran Ahmed M. Eltawil Raghu Rao* Chris Dick* University of California, Irvine *Xilinx Inc. ABSTRACT In this paper, we present the architecture of a highly configurable multi-input multi–output (MIMO) orthogonal frequency division multiple access (OFDMA) platform. The platform is designed to support experimentation with various communication algorithms, thus allowing an intimate understanding of the performance of complex algorithms under real-life constraints. The hardware used is the wireless open access research platform (WARP) which facilitates rapid prototyping utilizing the FPGA and multi radio interfaces available. Categories and Subject Descriptors B.0 [Hardware]: GENERAL General Terms Algorithms, Performance, Design, Experimentation, Verification. Keywords: MIMO, OFDMA, WiMax, FPGA, Wireless. 1. INTRODUCTION Multi Input Multi Output (MIMO) concepts coupled with Orthogonal Frequency Division Multiple Access (OFDMA) has attracted significant attention in recent years by providing high data rates, more reliable links and potentially higher network capacity. However, accurately quantifying the performance of such systems is non-trivial due to the fact that much of the favorable characteristics of MIMO techniques are highly dependent on the nature of the wireless channel. To gain performance benefits, advanced standards utilize a multitude of modalities that aim to maximize throughput based on current channel conditions. Thus, to fully characterize the performance of a system under all possible channel variations and modalities requires a vast sample space of simulation channels. These demanding guidelines require experimental platforms to perform accurate real time evaluation and confirmation of theoretical assumptions. This paper presents the architecture and design of a downlink transceiver for WiMax systems utilizing a reconfigurable Xilinx FPGA based board (WARP board) [1]. Section 2 presents the system architecture while section 3 discusses the hardware platform. Section 4 describes the demonstration procedure. The paper is concluded in section 5. 2. System Architecture Base station (BS) and mobile station (MS) in WiMax exchange information within frames. A WiMax frame is divided in two parts, the downlink portion and the uplink portion. In the downlink portion (focus of this paper), the BS establishes the link by transmitting a preamble for timing and initial frequency synchronization followed by link parameters FCH, DL-MAP and UL-MAP as explained in [2]. The pay load of the downlink sub- frame follows and contains information destined to multiple MSs. At the MS end, first, the preamble is detected, then link parameters are decoded and finally the information that is addressed to it is extracted from the payload. Figure 1 and Figure 2 illustrate the building blocks of the BS transmitter and the MS receiver respectively to carry out this process. Each block is designed with parameters as specified in the standard (IEEE802.16e [2]). Table 1 summarizes these parameters. 3. HARDWARE PLATFORM The WARP boards developed by Rice University [1] were chosen as an ideal platform for rapid prototyping. This platform had two radio boards connected to a Xilinx Virtex II FPGA. The transmitter node receives the data to send over the air from a PC. The source of data could be a picture or a video. At the receiver end the processed data is sent to another PC to be reconstructed as a picture or video depending on the data source. Communications between the PCs and the boards are performed through Ethernet ports and according to TCP/IP protocols. Figure 3 illustrates the link described above. 4. Demonstration Description The transmitter module will be running in real time on the WARP boards, however the receiver module will be running in a non-real time mode. Samples received by the RF front-end are sampled by the analog to digital converters and stored as files on the host receive PC. These input files can then be further processed by MATLAB on the PC to evaluate a host of different algorithms. Examples of algorithms currently being investigated (and reported on) within the Wireless Systems and Circuits group at the University of California, Irvine include a) Robust channel estimation techniques and power loading techniques [3], b) Sphere decoding for MIMO systems [4,5], c) MIMO OFDMA relay networks [6,7] among others. The demonstration proceeds as follows: 1. On the transmitter PC and in MATLAB one picture per MS is converted to bits. 2. The bits are sent to the transmit WARP board to be modulated, coded and put in WiMax frames and broadcast. 3. Receiver starts sampling and storing the data upon detection of a packet. 4. The received and sampled data is then transferred to the receiver PC and saved in a file. 5. The file is processed and the information destined to the MS of interest is extracted and decoded. Copyright is held by the author/owner(s). MobiHoc’09, May 18–21, 2009, New Orleans, Louisiana, USA. ACM 978-1-60558-531-4/09/05. 325