The motivation of this thesis is to contribute to the improvement of the physical-layer secrecy and privacy of wireless communication. Firstly, the rate and power adaptation technique is investigated to improve the energy efficiency of the physical-layer secrecy. We present the optimum rate and power adaptation rule that maximizes the average secrecy energy efficiency(SEE) subject to an average transmission power constraint. The SEE is defined as the outage secrecy capacity, the largest secrecy rate, such that the outage probability is less than a certain value, divided by the total power consumption (bits per joule). We also characterize the SEE gain provided by varying the rate and/or the power, and discuss the impact of the number of antennas on the optimum adaptation rule. Secondly, the joint impact of imperfect knowledge of the channel gain (channel uncertainty) and noise power (noise uncertainty) at the adversary is investigated to improve the physical-layer privacy. We characterize the covert throughput gain provided by the channel uncertainty as well as the covert throughput loss caused by the channel fading as a function of the noise uncertainty. We also show the impact the channel uncertainty on the total detection error probability and the covert throughput. Our result shows that the channel fading is crucial to hiding the signal transmission, particularly when the noise uncertainty is low and/or the receive SNR is high. The impact of the channel uncertainty on the total detection error probability and the covert throughput is more significant when the noise uncertainty is larger. Finally, hiding a covert (private) message in non-orthogonal multiple access (NOMA) systems by superimposing (embedding) it under other messages is proposed.We determine the total detection error probability (sum of false alarm and missed detection probability), the adversary's optimum detection strategy that minimizes the total detection error probability, and the communicator's optimum message hiding strategy that maximizes the total detection error probability. Additionally, we explore exploiting the channel variations to further increase the total detection error probability. We show that the total detection error probability increases and converges to 1 as the number of users increases and that the total detection error probability, hence the covert rate, can be increased by increasing the transmission power when the channel variation is exploited.