ZP-CI/OFDM: A Power Efficient Wireless Transmission Technology

This work is supported by the National Natural Science Foundation of China under Grant No. 61071102.

    Orthogonal Frequency Division Multiplexing (OFDM) is based on Multi-Carrier Modulation (MCM).  In OFDM, the subcarriers are overlapped but do not affect each other so that the data can be efficiently transmitted based on frequency division multiplexing. This technology has many advantages: Intersymbol Interference (ISI) is countered, spectum can be efficiently utilized, and data can be transmitted at high speed. As a result, it has attracted much attention [1]. 

    In the 1980s, OFDM began to be applied in the telecommunications industry. In the 1990s, it was widely applied in video and audio broadcasting services such as Asymmetric Digital Subscriber Line (ADSL), Very High bit rate Digital Subscriber Line (VHDSL), Digital Audio Broadcasting (DAB), and Digital Video Broadcasting (DVB) [2]. In 1999, IEEE approved the 802.11a standard for 5 GHz wireless local area networks [3], employing OFDM-based physical layer transmission. In the IEEE 802.16 standard [4], OFDM is treated as a basic technology for the physical layer. In a Long Term Evolution (LTE) system, Orthogonal Frequency-Division Multiple Access (OFDMA) is used on the forward link, and Carrier Frequency Division Multiple Access (SC-FDMA) is used on the reverse link [5]. The IEEE 802.15.3a standard for short-distance communication also treats OFDM as an alternative to Ultra-Wideband (UWB) technology [6]. In short, OFDM has become a mainstream transmission technology in broadband wireless communications.

    However, OFDM has some defects: 
    (1) An OFDM system outputs the combined signals of multiple independent subcarriers, and the synthesized signal produces much higher Peak-to-Average Power Ratio (PAPR) compared to signals of single-carrier systems. High PAPR puts high requirements on the linearity of the Power Amplifier (PA) of the transmitter and reduces the transmitter’s power efficiency.

    (2) By changing the frequency selective fading channel to flat fading sub-channels, an OFDM system can counter ISI and simplify equalization processing at the receiver. However, OFDM is deprived of frequency (multipath) diversity gain when deep fading occurs in some subcarriers because data symbols on these subcarrier are extremely difficult to detect. This limits Bit Error Rate (BER) performance of the OFDM system and reduces power efficiency.
To solve the problem of low power efficiency in traditional OFDM systems, Wiegandt et al. introduced Carrier Interferometry (CI) codes into the OFDM system and proposed an enhanced OFDM transmission technology called CI/OFDM [7],[8]. Unlike traditional OFDM systems,
CI/OFDM systems do not transmit low-speed concurrent data via respective subcarriers but employs orthogonal CI codes to spread the data over all subcarriers for concurrent transmission. Hence, CI/OFDM brings about frequency diversity gain and enhances the system’s BER performance without sacrificing transmission rate. Moreover, in the time domain, CI codes enable the peaks of time-domain waveforms to be evenly staggered. This is unlike OFDM systems, where the output comprises the addition of many random sine signals. Thus the problem of high PAPR is eliminated.

    Traditional OFDM eliminates ISI by using Cyclic Prefix (CP) as a guard interval. Recent research shows that an OFDM system using Zero-Padding (ZP), that is ZP-OFDM, can recover transmission symbols in the case of deep fading channel. Therefore, ZP-OFDM has better BER performance than traditional CP-based OFDM [9].

1 System Model
    Fig. 1 shows the ZP-CI/OFDM system model. At the transmitter, Inverse Discrete Fourier Transform (IDFT) is used for spreading CI code [10]. Data is modulated via an N-point IDFT onto each subcarrier, and Ng zeros are added at the end of the data symbol as a guard interval for realizing ZP-OFDM-based transmission. At the receiver, the signals can be detected in either frequency or time domain, thus frequency diversity gain can be fully used and the system’s power efficiency can be improved.

2 Receiver Technologies
    In a ZP-CI/OFDM system, signal detection technology at the receiver is critical for power efficiency. Three signal reception models are represented by key signal detection technologies: Frequency-domain Minimum Mean Square Error (MMSE) detection, time-domain MMSE detection, and nonlinear detection.

2.1 Frequency-Domain MMSE Detection
    Frequency-Domain MMSE (FDMMSE) detection is designed for the frequency-domain signal reception model. The receiver converts received time domain symbols into frequency domain symbols via an (N+Ng)-point Discrete Fourier Transform (DFT). It then estimates a frequency domain channel matrix H of order (N+Ng), which is a diagonal matrix 
H=diag(H0,H1,…,HN+Ng -1)  (1)
H0,H1,…,HN+Ng -1 =
FN+Ng (h0,…,hL,0,…,0)(N+Ng)×1   (2)

    where FN+Ng denotes DFT matrix of order (N+Ng), and (h0,…,hL) is the Channel Impulse Response (CIR) vector of fading channel. The receiver uses the matrix H to conduct MMSE detection of frequency domain signals and performs DFT to de-spread CI codes in order to recover the original signals. The schematic diagram of FDMMSE detection is shown in Fig. 2.

2.2 Time-Domain MMSE Detection
    As the FDMMSE detection algorithm cannot make full use of the system’s frequency diversity gain, a Time-Domain MMSE (TDMMSE) detection algorithm is further proposed. This technology is based on the time domain signal reception model. By means of time domain channel estimation, the receiver estimates a (N+Ng) × N time domain channel matrix h, which is a truncated rectangle Toepitz matrix. Then, the matrix h is used to conduct MMSE detection of received time domain signals and performs DFT to convert the signals from time domain to frequency domain. Finally, the receiver performs DFT to de-spread CI codes in order to recover the original signals. The schematic diagram of TDMMSE detection is shown in Fig. 3.

2.3 Nonlinear Detection
    To further make use of diversity gain and improve power efficiency, ZP-CI/OFDM can adopt a nonlinear detection algorithm that is more complicated than time and frequency domain MMSE detection algorithms. This detection method is based on the prerequisite that the signal reception model can be equivalent to an N ×(N+Ng) Multiple Input and Multiple Output (MIMO) system. On this basis, some nonlinear MIMO detection algorithms can be used for signal detection, thus improving system performance.

    The receiver estimates an N ×(N+Ng) time domain channel matrix h. Then it analyzes received signals and generates an equivalent N ×(N+Ng) MIMO matrix Ω, which is from digitally modulated data symbols to received signals,
Ω = hFN-1FN-1     (3)

    where FN-1 is an IDFT matrix of order N. Finally, the receiver uses existing nonlinear detection algorithms such as Ordered Successive Interference Cancellation (OSIC) [11] and Sphere Decoding (SD) [12], for nonlinear detection of the received signals and recovery of the original signals. The schematic diagram of nonlinear detection technology is shown in Fig. 4.

3 Simulation
    To verify the high power efficiency of a ZP-CI/OFDM system, simulation tests were conducted to measure the Bit Error Rate (BER) and Peak-to-Average Power Ratio (PAPR) by comparing with traditional CI/OFDM and OFDM.

3.1 BER
    Simulation tests were conducted under the following conditions: The channel model was COST207TUx6 [13], the modulation scheme was 16-State Quadrature Amplitude Modulation (16-QAM), the bandwidth was 2.5 MHz, the number of subcarriers was 128, the length of guard interval was 16, and the maximum Doppler shift was 40 Hz.

    The simulation results (Fig. 5) demonstrate that BER of ZP-CI/OFDM system is lower than that of CI/OFDM and OFDM systems regardless of which detection technology (FDMMSE, TDMMSE, or OSMMSE) is used. The higher the Signal-to-Noise Ratio (SNR), the more obvious BER gain is. With better use of frequency diversity gain, a ZP-CI/OFDM system uses power more efficiently than an OFDM or CI/OFDM system.

3.2 PAPR
    In PAPR simulation tests, Complementary Cumulative Distribution Function (CCDF) is introduced to describe the signal’s PAPR. Fig. 6 shows the simulation results of PAPRs of ZP-CI/OFDM, CI/OFDM, and OFDM signals modulated with 16-QAM. CI-OFDM PAPR is the lowest owing to the introduction of CI code. In a ZP-CI/OFDM system, PAPR is slightly higher than that of CI-OFDM because ZP is employed as a guard interval method. However, PAPR is much lower than that of an OFDM system. Therefore, in a ZP-CI/OFDM system, much more BER gain is obtained at the cost of small PAPR.

4 Conclusion
    This paper analyzes the advantages and disadvantages of OFDM technology, regarded as a mainstream transmission technology for broadband wireless communications. Low power efficiency is one of the deficiencies that hinder the development of OFDM based wireless transmission technologies. To solve the problem, ZP-CI/OFDM is proposed in this paper. By spreading transmission symbols to all OFDM subcarriers via CI codes, PAPR is reduced, and frequency diversity gain is exploited for ZP-CI/OFDM system. Moreover, by zero-padding at the transmitter, ZP-CI/OFDM can adopt some advanced receiver technologies to make better use of frequency diversity gain and to improve the system’s power efficiency.

References
[1] R. Van Nee and R. Prasad, OFDM for wireless multimedia communications. London: Artech House, 2000.
[2] H. Rohling, T. May, K. Bruninghaus, and R. Grunheid, “Broadband OFDM radio transmission for multimedia applications,” Proc. IEEE, vol. 87, no. 10, pp.1778-1789, 1999.
[3] Draft Supplement to Standard for Telecommunications and Information Exchange between Systems - LAN/MSN Specific Requirements - Part 11: Wireless MAC and PHY Specifications: High Speed Physical Layer in the 5 GHz band, IEEE P802.11a/D6.0, 1999.
[4] Air Interface for Fixed and Mobile Broadband Wireless Access Systems, IEEE 802.16e, 2006.
[5] Proposed Study Item on Evolved UTRA and UTRAN, 3GPP RP-040461, 2004.
[6] TI Physical Layer Proposal: Time-Frequency Interleaved OFDM, IEEE P802.15-03/141r3-TG3a, 2003.
[7] D. Wiegandt, C. Nassar, W. Zhiqiang.“High-throughput, high-performance OFDM via pseudo-orthogonal carrier interferometry spreading codes”, IEEE Trans. Commun., vol. 51,    no. 7., pp.1123-1134, 2003.
[8] W. Zhiqiang, W. Zhijin, D. Wiegandt et al. “High performance 64-QAM OFDM via carrier interferometry spreading codes,” Proc. 58th Vehicular Technology Conf. (VTC-Fall’03), Orlando, FL, 2003, vol.4, pp. 557-561.
[9] B. Muquet, Z. Wang, G. Giannakis et al. “Cyclic prefixing or zero padding for wireless multicarrier transmissions,” IEEE Trans. Commun., vol. 50, no. 12, pp. 2136-2148, 2002.
[10] F. Xu, R. Xu, and Haixin Sun. “Implementation of carrier interferometry OFDM by using pulse shaping technique in frequency domain,” Proc. 2007 IEEE Int. Workshop on Anti-counterfeiting, Security, and Identification, Xiamen, China, pp. 319-323.
[11] M. Debbah, B. Muquet, M. De Courville et al. “A MMSE successive inteference cancellation scheme for a new adjustable hybrid spread OFDM system,” Proc. 51st Vehicular Technology Conf. (VTC-Spring’00), Tokyo, 2000, vol. 2, pp.745-749.
[12] E. Agrelle, T. Eriksson, A. VARDY et al. “Closet point search in lattices,” IEEE Trans. Inf. Theory, vol. 48, no. 8, 2002, pp.2201-2214.
[13] “COST 207: Digital Land Mobile Radio Communications,” Office for Official Publications of the European Communities, Luxembourg, 1989.

Biographies

Pei Gao (gaopei@uestc.edu.cn) is working towards an M.S. degree at the National Key Laboratory of Science and Technology on Communications, University of Electronic Science and Technology of China. His research interest is signal processing technologies for wireless communications. He has participated in three funded projects, published two papers indexed by EI, and has applied for one Chinese patent.

Xiaohu Chen (cxh4389@163.com) is working towards an M.S. degree at the National Key Laboratory of Science and Technology on Communications, University of Electronic Science and Technology of China. His research interests include signal processing technologies and waveform design for wireless communications. He has participated in three funded projects and applied for one Chinese patent.

Jun Wang (junwang@uestc.edu.cn) is an associate professor at the National Key Laboratory of Science and Technology on Communications, University of Electronic Science and Technology of China. His research interest is signal processing technologies for wireless communications. He has participated in eight funded projects, published 60 papers indexed by SCI/EI, and applied for eight Chinese patents.

[Abstract] Low power efficiency is a deficiency in traditional Orthogonal Frequency Division Multiplexing (OFDM) systems. To counter this problem, a new wireless transmission technology based on Zero-Padding Carrier Interferometry OFDM
(ZP-CI/OFDM) is proposed. In a ZP-CI/OFDM system, transmission symbols are spread to all OFDM subcarriers via carrier interferometry codes. This reduces the Peak-to-Average Power Ratio (PAPR) that traditional OFDM suffers and also exploits frequency diversity gain. By zero-padding at the transmitter, advanced receiver technologies can be adopted for ZP-CI/OFDM so that frequency diversity gain can be further utilized and the power efficiency of the system is improved.

[Keywords] power efficiency; carrier interferometry; orthogonal frequency division multiplex; zero-padding; frequency diversity gain