Mobility Adaptation For OFDM Systems Under Fast Fading Channels

By the end of 2015, China's operational high-speed railway (HSR) mileage reached 19,000 kilometers, occupying 60% of the whole world's HSR operational mileage. On the other hand. China has also made significant breakthrough in the field of wireless com-munications. Time duplex-long term evolution (TD-LTE), the fourth generation (4G) of mobile communications system proposed by China has entered into the large-scale com-mercial stage. Meanwhile, with the rapid development of high-speed railway, there are higher requirements for the train dispatching, automatic train control, and broadband wireless communication services for large numbers of passengers in high-speed train. Un-der high mobility scenarios, Doppler effect will lead to fast fading for wireless channels and highly sealed carriages will lead to signal loss seriously. All of these will affect system performance dramatically. In this thesis, the adaptive transmission technique is used to counteract the deterioration of system performance. Three kinds of adaptation schemes will be studied. The first one is traditional adaptation scheme based on channel state information (CSI) under time-varying fading channels. The second one is mobility adap-tation based on velocity information only under rapidly time-varying fading channels. The third one is an improved mobility adaptation scheme based on velocity and channel under doubly-selective fading channels. As it is very hard to get the accurate CSI under fast fading channels for traditional adaptation scheme and improved mobility adaptation scheme, the imperfect CSI is considered in the thesis.Firstly, the traditional adaptation scheme based on CSI is usually employed to adap-tively adjust the system parameters. This thesis proposes three kinds of CSI based adap-tive schemes. The first one is a new conditional power and rate adaptation scheme for orthogonal frequency-division multiplexing (OFDM) systems in the presence of carrier frequency offset (CFO), with perfect and imperfect CSI. The optimal power adaptation is to maximize the spectral efficiency of an OFDM system using M-ary quadrature ampli-tude modulation (MQAM) under average power and instantaneous bit error rate (BER) constraints. Closed-form expressions for the average spectral efficiency (ASE) of adaptive OFDM systems are derived for perfect and imperfect CSI cases. Aimed at maximizing the ASE, the second one is a power and rate adaptation scheme for OFDM systems in the presence of very rapidly time-varying fading channels, under instantaneous BER con-straints for each subcarrier. The closed-form expressions for the ASE of adaptive systems with perfect and imperfect CSI are obtained. The impact channel estimation error and the delay are considered. Also aimed at maximizing the ASE, the third one is a rate and subcarrier bandwidth adaptation (RSBA) scheme for OFDM systems in the presence of frequency selective and very rapidly time-varying fading channels. The closed form expressions for rate and capacity with perfect CSI and imperfect CSI are derived. The imperfect CSI is caused jointly by channel estimation error and unavoidable delay which results outdated CSI. Schemes mentioned above maximize ASE and reduce the impact of the channel estimation errors and delay under high mobility scenarios.Secondly, under rapidly time-varying fading channel, it is very hard to obtain accurate instantaneous CSI because of channel estimation error and unavoidable delay. However, obtaining the velocity of the moving terminals is not difficult, where the velocity (mobility) can be considered as a large scale metric. A new concept called mobility adaptation transmission over very fast fading channels is proposed, which adjusts system parameters based on the geographic information of the terminals instead of CSI at the transmitter. Two schemes are proposed based on mobility adaptation. The first one is the optimal bit and rate adaptation based on velocity only. The results show that the proposed scheme could improve the system performance significantly. In the second scheme, a prior knowledge of velocity, such as velocity distribution, at the transmitter is employed in the mobility adaptation. The power and rate adaptation based on velocity only is proposed to maximize the ASE. The results show that mobility adaptation with a prior knowledge of velocity at transmitter is better than that of without a prior knowledge of velocity at transmitter. Especially, the improvement of performance increases as the variance increases due to larger velocity variation.In addition, improved mobility adaptation based on the terminal velocity and CSI at the transmitter is proposed. In practical cellular systems, channel estimation is normally performed, even if it may not be accurate. Therefore, the inaccurate CSI can still be used in conjunction with velocity information to improve adaptive transmission performance. Under the intermediate velocity scenario, a power and rate adaptation scheme for OFDM systems, based on perfect and CSI and velocity joint is presented. The proposed mobility adaptation scheme can maximize the ASE under average power and instantaneous BER constraints. The imperfect CSI is caused jointly by channel estimation error based on minimum mean square error (MMSE) detection criteria at the receiver and unavoidable delay. It is shown by simulation results that the improvement of the joint adaptation scheme based on velocity and CSI, compared with traditional adaptation scheme based on CSI and mobility adaptation scheme based on velocity, is quite significant.Finally, for variable mobility scenarios, an integrated mobility adaptation scheme is proposed. Under the stationary or low mobility, the traditional adaptation scheme should be adopted because the CSI could be obtained accurately. If the velocity of terminal is intermediate, the best choice is the improved mobility adaptation based on velocity and CSI because the imperfect CSI is also useful to improve system performance. However, for the high mobility, the best choice is mobility adaptation based on velocity only because estimated CSI is so poor and cannot be used due to severe channel estimation error and outdated time delay.