Youn Ok Park

Very fast detection for rate-2 quasi-orthogonal STBCs

For an important and standard rate-two quasi-orthogonal space-time block code, we propose a very fast detection scheme which achieves a wide range of tradeoff between performance and complexity. In the proposed scheme, exploiting the special quasi-orthogonal structure of the code, we efficiently obtain multiple candidate constellation points around the maximum likelihood (ML) point. Furthermore, we develop a very fast sorting method for the multiple candidate points to further improve the detection speed. Numerical results demonstrate that the proposed scheme can perform very close to the maximum likelihood detection at an extremely low complexity.

Design and analysis of a smart antenna system for IMT-2000 W-CDMA

In this paper, we design and analyse the performance of an IMT-2000 wideband code division multiple access (W-CDMA) smart antenna system with 5 MHz bandwidth in the reverse link. Dedicated pilot channel and 9.8 kbps traffic channel data are included in our simulation. The dedicated pilot channel is used not only to compensate channel error but to estimate cross-correlation vector for beamforming. Simulation results are given in the form of bit error rate (BER) vs. E/sub b//I/sub o/ under various parameters including mobile speed, power control error, and the number of multipaths in Rayleigh fading. On average, 5.4 dB E/sub b//I/sub o/ gain is attained when the number of available paths is increased from one to two in the case of BER 10/sup -3/. The effect of power control error is so severe that 3 dB E/sub b//I/sub o/ gain is required when the power control error increases from 10% to 20%. However, the effect of mobile speed on BER is less than 1.6 dB at most.

Design of FFT processor for IEEE802.16m MIMO-OFDM systems

In this paper, an area-efficient FFT processor is proposed for IEEE 802.16m mobile WiMAX systems. The proposed scalable FFT processor can support the variable length of 512, 1024, 2048 and 4096. By reducing the required number of non-trivial multipliers with mixed-radix (MR) and multi-path delay commutator (MDC) architecture, the complexity of the proposed FFT processor is dramatically decreased without sacrificing system throughput. The proposed FFT processor was designed in hardware description language (HDL) and synthesized to gate-level circuits using 0.18um CMOS standard cell library. With the proposed architecture, the gate count for the processor is 49K and the size of memory is 96Kbits, which are reduced by 12% and 26%, respectively, compared with those of the 4-channel radix-2 MDC (R2MDC) FFT processor.

BER performance of AGC in high-speed portable Internet system

The paper proposes a practical method for AGC (automatic gain control) in an HPI (high-speed portable Internet) system. Downlink packets in this system consist of preamble symbols for AGC, AFC (automatic frequency control) and other purposes, and data symbols for traffic transmission. We compare the theoretical BER performance with the simulation one and produce an optimum parameter for AGC in this system. The proposed AGC method utilizes the preamble and requires a small amount of hardware. The validity of the proposed method is verified by computer simulation and real implementation.

Fast ML Detection for Quasi-Orthogonal STBCs of Rate 2 in Under-Determined Systems [Transactions Papers]

A fast maximum-likelihood (ML) detection scheme is proposed for the important quasi-orthogonal space-time block codes (STBCs) of symbol rate 2 for three and four transmission antennas. In particular, the proposed scheme is applicable even to the under-determined system where the number of transmission antennas is larger than that of the receive antennas. The half of the symbols of the transmission matrix are detected by two dependent double-layer sphere decoders, which are referred to as dual double-layer sphere decoding in this paper. The remaining half of the symbols are detected by single-symbol detection, of which complexity is extremely low. Numerical results demonstrate that the proposed new detection method improves the detection speed up to ten times depending on the signal- to-noise ratio (SNR) compared to the state-of-the-art scheme.

LS channel estimation performance analysis for RoF channel environment in the OFDM system

Generally in real implementation of OFDM system, the simple and fast LS channel estimation method has been used for the wireless channel in the OFDM systems. If the wireless channel environment were mixed with RoF (Radio Over Fiber) channel, could we still use the conventional least square (LS) channel estimation method for proper communication? Therefore, in this paper, we analyzed that the performance of LS channel estimation where both wireless and RoF channel are coexisted in the OFDM based mobile communication system.

A traffic-efficient fronthaul for the cloud-RAN.

As the mobile communication technology evolves towards the cloud RAN based 4G and 5G systems, data traffic of a BS fronthaul which is mainly packed up of sampled baseband TRx signals increases geometrically and makes it hard the existing technologies over optical link handle such a traffic. In this paper, a new fronthaul based on RRPH which integrates baseband PHY into the RRH in order to minimize the fronthaul data traffic rate is proposed. Issues come from the architecture of new fronthaul based BS and the implementation results of it to support IEEE 802.16m specification are also introduced.

Diversity Analysis of Minimum Distance Based Relay Selection Schemes for Two-Way Relaying Systems with Physical Network Coding

A relay selection method is one of popular approaches to overcome fading effects in wireless relay channel due to its simplicity. However, previous works were concentrated on the case of analog network coding systems. In this paper, we consider a relay selection scheme in two-way relaying systems with physical network coding. Since conventional schemes based on analog network coding does not provide effective performance in physical network coded systems, we propose new selection criteria which maximize the minimum distance at the multiple access phase to achieve full diversity. Moreover, through asymptotic error rate analysis, it is confirmed that the proposed scheme is able to attain full diversity. Simulations results are also provided to verify the analysis of the proposed scheme.

RRH selection and UE transmission time adjustment for LTE-advanced uplink MU-MIMO in Distributed Antenna System environment

In this paper, after we investigate the influence of the arriving time difference on LTE-Advanced uplink MU-MIMO in Distributed Antenna System (DAS) environment, we suggest RRH selection and transmission time adjustment method for obtaining the receiving performance gain. DAS which is one of CoMP technologies has known as the most efficient technology for improving spectral efficiency and cell-edge user throughput. However, when uplink MU-MIMO is carried out in the DAS environment, the arriving time difference is inevitable owing to geographical distribution of RRHs location. The arriving time difference of the multiple UEs has decisive effect on system performance. Therefore, in this paper, we propose RRH selection and UE transmission time adjustment method for minimizing the arriving time difference. Through link-level simulation, we show the impact of the arriving time difference on LTE-Advanced uplink MU-MIMO in DAS environment and compare the performance of the suggested method with that of the conventional method.

Fast and robust timing synchronization for IEEE 802.16 MSs in mixed mode (16e, 16m)

When the ABS supports mixed mode that is it supports both 16e MSs and 16m AMSs, if we have a Software Defined Radio which supports both 16e and 16m modes, then we can take advantage of the extra preamble available from the other mode to do either fast or robust coarse timing synchronization. For 16e mode, if the signal is weak and cannot synchronize on 16e preamble, the MS can first synchronize on 16m PA-Preamble and then offset back by a known integer number of 16e symbols to reach 16e preamble location. Usually for a given performance or bit error rate, the 16m PA-Preamble can work at 5dB SNR lower than the 16e preamble. Similarly when working in 16m mode, if the AMS wants to do fast synchronization and the signal quality is good, it can synchronize on 16e preamble first (which is available every frame as compared to 16m PA-Preamble which is only available every super frame) and then offset forward to 16m A-Preamble (PA-Preamble or SA-Preamble). This way we can make choices based on SNR level and time required to synchronize.