Research and Simulation of Channel Estimation Based on IEEE802.16D
IEEE802.16d is a fixed broadband wireless access system that can provide up to 70 Mb / s peak transmission rate to support the deployment of different Qosl-type integrated data services. Firstly, the frame formation in Wireless MAN-0FDM based on IEEE802.16d is introduced. After analyzing the 0FDM channel estimation technique and interpolation algorithm, the estimated performance of the pilot and preamble in the protocol under the SUI channel model is simulated. The effects of different estimation algorithms and interpolation on the performance of the system are simulated. From the simulation results, it can be seen that the estimation effect of the preamble is better than the pilot, and the applicable conditions of the system pilot estimation are given.
Keywords: IEEE802.16d; OFDM; channel estimation; interpolation
0 Introduction Since the beginning of the 21st century, with the rapid development of the Internet and the increasing demand for various real-time multimedia services, broadband wireless technology will present huge development potential. The IEEE802.16d standard, as a fixed broadband wireless access solution for wireless metropolitan area networks (WMAN), has attracted much attention for its excellent performance and broad prospects. IEEE802.16 published the IEEE802.16d standard for fixed broadband wireless access systems in 2004. The physical layer of the standard defines four transmission modes, namely Wireless MAN-SC in the frequency range of 10 to 66 GHz, and three non-line-of-sight (NOLS) modes applied at 2 to 11 GHz: Wireless MAN-Sca , Wireless MAN—OFDM, Wireless—MAN—OFDMA. This article only discusses and analyzes WirelessMAN-OFDM. Since the wireless channel is not as fixed and predictable as the wired channel, the wireless OFDM communication system is seriously affected by the shadow fading and frequency selective fading of the wireless channel, so efforts must be made to reduce the influence of the wireless channel, which is the channel estimation of the wireless OFDM system. The technology presents a great challenge. The quality of channel estimation will directly affect the performance of the entire system. In this paper, based on IEEE802.16d, the preamble and pilot frequency are used as training symbols to simulate the performance estimation under the SUI channel model, and the simulation results are analyzed and studied.
1 Frame structure in Wireless MAN-OFDM Because OFDM modulation can effectively resist the multipath fading of wireless channels, it is used in the physical layer technology of Wireless MAN-OFDM and Wireless-OFDMA for NLOS applications below 11 GHz. The OFDM physical layer supports frame-based transmission. Figure 1 shows a schematic diagram of its downlink frame structure.
It can be seen from FIG. 1 that the communication data unit (PDU) of a downlink physical layer is composed of a preamble, a frame control header (FCH), and burst OFDM data. The preamble is mainly used to make various estimates. It consists of two consecutive special OFDM symbols. The first OFDM symbol uses only subcarriers whose sequence number is a multiple of 4, and its time-domain waveform includes four repeated 64 samples, CP in front. The second OFDM symbol uses only even-numbered subcarriers, and its time-domain waveform includes two repeated 128 samples, preceded by CP. Its time domain structure diagram is shown in Figure 2.
In the frequency domain, the first OFDM symbol frequency domain data is derived from the full-band prearnble multiple of 4 subcarrier data. The frequency domain of the 4th 64-sequence is defined as:
The frequency domain sequence of the full bandwidth preamble is given by the protocol. In burst OFDM data, the number of IFFT points for each OFDM symbol data is 256 points, that is, there are 256 subcarriers, which are divided into three types of subcarriers. The data subcarriers are used to transmit data. Pilot subcarrier (pi1ot), there is a pilot subcarrier every 25 data subcarriers, a total of 8, mainly used for various estimation. Empty subcarriers, that is, DC subcarriers and guard bands, do not transmit any data. The frequency domain structure of the OFDM symbol is shown in Figure 3.
2 Channel estimation algorithm under IEEE802.16d Channel estimation is to estimate the frequency response of the wireless channel from the Transmitting Antenna to the receiving antenna. According to the received channel response, amplitude and phase distortions are generated and white Gaussian noise is added to the received sequence to accurately identify the transmission characteristics of the channel in the time or frequency domain. There are three types of channel estimation algorithms commonly used in OFDM systems based on pilot symbols and interpolation techniques as well as decision feedback and blind channel estimation. The analysis in this paper is based on the pilot symbol channel estimation under the IEEE 802.16d system. The principle of this kind of algorithm is to use the information known by the receiver to perform channel estimation. There are two types of pilot insertion methods: block-type pilot and block-type pilot. It is not difficult to find that in IEEE802.16d, both the pilot in each data OFDM symbol and the preamble before a frame are inserted in the form of a comb. Commonly used channel estimation algorithms are LS algorithm based on minimum variance criterion and LMMSE algorithm based on minimum mean square error criterion.
2.1 LS algorithm If it is assumed that H is the frequency domain response vector of the channel, X and Y are the frequency domain representation of the transmitted and received signal vectors respectively, and n is Gaussian white noise, then Y = XH + n. The LS algorithm is to minimize the square error of equation (1):
It can be seen from the above formula that the LS algorithm is greatly affected by noise.
2.2 LMMSE algorithm The LMMSE algorithm is to make the mean square error of the formula HLMMSE = argminE [(HLMMSE-H) (HLMMSE-H) H] the smallest. The LMMSE algorithm can be obtained on the basis of the LS algorithm:
In equation (2), RHH = E [HHH] is the autocorrelation matrix of the channel impulse response, which can be obtained according to the statistical characteristics of the channel. σ2n is the variance of additive Gaussian noise.
2.3 Interpolation algorithm Under IEEE802.16d, no matter whether it is based on pilot symbols or based on preamble channel estimation, LS and LMMSE algorithms have the problem of interpolation of channel estimation, that is, the channel response value of the non-pilot point It can only be obtained by interpolating the channel response values ​​of the pilot points. This paper mainly applies and compares the performance of simple linear interpolation and two-dimensional linear interpolation.
2.3.1 Linear interpolation Linear interpolation is the simplest algorithm in the interpolation algorithm. The channel response value at the non-pilot point can be expressed as follows:
Get, where mL≤k
Among them: C0 = one (α-1) (α + 1); Cl = α (α-1) / 2
α = l / N. Two-dimensional interpolation reduces interpolation errors and can obtain better performance.
3 Simulation parameter settings and result analysis The simulation platform here is based on the downlink end of the IEEE 802.16d system. The OFDM symbol parameters and system channel bandwidth are shown in Table 1, respectively.
The simulated frame length is 5 ms, and one frame contains 69 OFDM symbols. The system uses 16QAM modulation without considering RS-CC coding and interleaving in the protocol. The construction of the entire system and the simulation of the channel estimation algorithm are realized through M files in Matlab. The simulation results are based on the system's bit error rate (BER) as its performance standard. The obtained simulation results are shown in Figures 4 to 7 respectively.
It can be easily seen from Figure 4 that Gauss interpolation is better than linear interpolation, so the subsequent simulations all use Gauss interpolation to improve system performance. Figure 5 is the performance simulation diagram of LS and LMMSE under the SUI3 channel under the preamble estimation. It can be seen from the figure that the performance of the LMMSE algorithm is better than LS, and its performance is improved by about 3 dB compared with the LS algorithm, which reflects the LS The algorithm is susceptible to noise, but at the same time, the operation complexity of LMMSE is greater than that of LS. Figure 6 shows the LS estimation performance of the pilot and preamble under SUI3. From this figure, it can be clearly seen that the pilot estimation performance requires pilot estimation. When the signal-to-noise ratio is lower than 15 dB, the performance of the two The performance gap is not large, but this gap becomes more obvious as the signal-to-noise ratio increases. This is mainly because SUI3 is a slow fading channel, and no pilot is used for channel tracking, and because the pilot point in the preamble is much larger than the pilot point in the OFDM symbol, the natural preamble estimation performance is superior to the pilot. When the channel is SUI4, it can be seen from Figure 7 that the performance difference between the two is even greater, and even then the pilot estimation is completely unavailable, far from meeting the performance requirements specified in IEEE802.16d, so this time The system can only use leading estimates. Analyze the reason and find that the maximum multipath delay under the SUI4 channel is 4μs, then the coherent bandwidth of the system B = O. 25 MHz, and the pilot interval in the OFDM symbol is 0. At 375 MHz, it can be seen that the interval between pilots is greater than the coherent bandwidth of the system, which leads to the deterioration of pilot estimation performance. Further, it can be concluded that the applicable condition of the system pilot estimation is that the system pilot interval must not be greater than the system bandwidth, otherwise the pilot estimation is undesirable.
4 Conclusion IEEE802.16d fixed wireless broadband access system has great advantages in terms of transmission speed, network construction distance and cost investment. It is the ideal solution for the current development and promotion of wireless access. Of course, as a new technology, there must be many difficulties to be solved, and channel estimation technology is one of the important aspects. This paper simulates the channel estimation performance based on the system under the SUI channel model. The simulation results show that the performance of the preamble estimation is better than that of the pilot. In SUI4, only the preamble estimation can be considered. The reason for the performance difference is also theoretically analyzed, and the applicable conditions of the pilot estimation are given. Some of the conclusions and simulation results are Further research on IEEE802.16d system has great reference significance.
Ac To Dc Transformer,Ef High Frequency Transformer,Rm 10 Electric Transfomer,110 Volt Transformer
IHUA INDUSTRIES CO.,LTD. , https://www.ihua-magnetics.com