Adaptive communication is one of the hottest areas of research in telecommunication engineering. This technique is recommended for many 3rd Generation (3G) and 4th Generation (4G) communication standards like WIFI (IEEE 802.11n/b/g) and WiMAX (IEEE 802.16/e) etc. These systems are mainly Orthogonal Frequency Division Multiplexing (OFDM) based communication systems. In OFDM systems there are a number of subcarriers that may exhibit different channel state information. In adaptive communication, the transmission parameters like code rate, modulation scheme and power are adapted with respect to the channel state information at different subcarriers, so that the overall throughput of the system may be maximized while satisfying certain constraints like bit error rate and power, at the same time. In this dissertation this problem is formulated in a unique way and three approaches are proposed to solve it. In first approach, we have proposed a fuzzy rule base system (FRBS) for adapting code rate and modulation scheme according to the channel state information (CSI) and quality of service (QoS) demands at individual subcarriers, respectively. QoS represents the required target bit error rate (BER) at any subcarrier. Power distribution was considered flat for all the subcarriers. FRBS is designed by incorporating sufficient number of rules in the rule base. A large portion of these rules is obtained by analyzing different code rates and modulation schemes while remaining portion of rules is added from expert knowledge. Feed forward convolutional codes of various rates, with constraint length 3 and Quadrature Amplitude Modulation (QAM) family are used as forward error correction and modulation scheme, respectively. In second approach, FRBS in contrast to various algorithms is proposed for adaptive coding, modulation and power for OFDM systems. In this scheme, FRBS takes care of adapting code rate and modulation symbol while other algorithms are used to adapt the transmit power. These algorithms include famous Water-filling algorithm, Genetic Algorithm and Differential Evolution Algorithm. In third approach, a real time adaptive coding, modulation and power scheme is proposed using fuzzy rule base system and Gaussian Radial Basis Function Neural Network (GRBS-NN). In this technique neural network was trained offline. Once trained, GRBS-NN immediately suggests the optimum transmit power vector, by feeding in the channel state information and QoS demands from all subcarriers. FRBS is used to choose optimum modulation code pairs for all the subcarriers. The proposed schemes are compared with well known techniques in the literature for adaptive communication in OFDM systems. Simulation results show the supremacy of proposed schemes over the other.