The scarcity of bandwidth has always been the main obstacle for providing reliable high data-rate wireless links, which are in great demand to accommodate nowadays and immediate future wireless applications. In addition, recent reports have showed inefficient usage and under-utilization of the available bandwidth. Cognitive radio (CR) has recently emerged as a promising solution to enhance the spectrum utilization, where it offers the ability for unlicensed users to access  the licensed spectrum opportunistically. By allowing opportunistic spectrum access which is the main concept for the interweave network model, the overall spectrum utilization can be improved. This requires cognitive radio networks (CRNs) to consider the spectrum sensing and monitoring as an essential enabling process for the interweave network model. 

Machine-to-machine (M2M) communication, which is the basic enabler for the Internet-of-Things (IoT), has emerged to be a key element in future networks. Machines are expected to communicate with each other exchanging information and data without human intervention. The ultimate objective of M2M communications is to construct comprehensive connections among all machines distributed over an extensive coverage area. Due to the radical change in the number of users, the network has to carefully utilize the available resources in order to maintain reasonable quality-of-service (QoS). Generally, one of the most important resources in wireless communications is the frequency spectrum. To utilize the frequency spectrum in IoT environment, it can be argued that cognitive radio concept is a possible solution from the cost and performance perspectives. Thus, supporting numerous number of machines is possible by employing dual-mode base stations which can apply cognitive radio concept in addition to the legacy licensed frequency assignment. 

In this thesis, a detailed review of the state of the art related to the application of spectrum sensing in CR communications is considered. We present the latest advances related to the implementation of the legacy spectrum sensing approaches. We also address the implementation challenges for cognitive radios in the direction of spectrum sensing and monitoring. We propose a novel algorithm to solve the reduced throughput issue due to the scheduled spectrum sensing and monitoring. Further, two new architectures are considered to significantly reduce the power consumption required by the CR to enable wideband sensing. Both systems rely on the 1-bit quantization at the receiver side. The system performance is analytically investigated and simulated. Also, complexity and power consumption are investigated and studied.

Furthermore, we address the challenges that are expected from the next generation M2M network as an integral part of the future IoT. This mainly includes the design of low-power low-cost machine with reduced bandwidth. The trade-off between cost, feasibility, and performance are also discussed. Because of the relaxation of the frequency and spatial diversities, in addition, to enabling the extended coverage mode, initial synchronization and cell search have new challenges for cellular-enabled M2M systems. We study conventional solutions with their pros and cons including timing acquisition, cell detection, and frequency offset estimation algorithms. We provide a technique to enhance the performance in the presence of the harsh detection environment for LTE-based machines. Furthermore, we present a frequency tracking algorithm for cellular M2M systems that utilizes the new repetitive feature of the broadcast channel symbols in next generation Long Term Evolution (LTE) systems. In the direction of narrowband IoT support, we propose a cell search and initial synchronization algorithm that utilizes the new set of narrowband synchronization signals. The proposed algorithms have been simulated at very low signal to noise ratios and in different fading environments.