Nowadays, life without wireless communications is inconceivable. Advances in CMOS technology and the creation of smaller and smaller circuits led to the use of wireless circuits in most today electronic devices. This progress also caused the development of micro-sensors. These micro-sensors have the ability to do numerous senses in things like voice recognition for sensing the earthquake. As well as, it has provided the collection of information in remote areas and places that are not suitable for human explorations. Therefore, recent advances in fabrication of small-scale integrated circuits in one hand and development of wireless communications technology on the other hand have created great motivation for researchers and engineers in the field of research and design of sensor networks. According to the development of wireless sensor networks, creation of simple and efficient communication with low power consumption and low cost between sensor nodes is an important and vital issue. In most of these types of networks, battery is used as an energy source. As changing the battery in a short time interval is impossible or costly in this type of networks, life of node comes to end with completion of battery energy. So in order to extend the lifetime of the network, it is required to manage the communication of nodes with each other and their energy consumption by using an appropriate protocol. That is why the subgroup 4 started working with similar purpose in Commission IEEE 802.15 that undertook the task of standardization in the field of personal wireless networks. Efforts of subgroup 4 from the Commission IEEE 802.15 led to provide IEEE 802.15.4 standard for two layers of PHY and MAC and the mentioned Union's efforts in 2004 led to provide a set of standards based on IEEE 802.15.4 standard with brand ZigBee . Because of the capabilities and benefits such as energy management, high throughput rates and scalability, IEEE 802.15.4 Protocol is a suitable options for wireless sensor networks. In this protocol, nodes compete with each other using the CSMA/CA algorithm in order to achieve the channel and sending their packs to destination. The algorithm consists of several different parameters that their regulation has a great impact on network throughput rate and nodes energy consumption. The aim of this thesis is to provide a solution to reduce the power consumption of ZigBee transmitter-receiver by improving the CSMA/CA algorithm that is used as node in wireless sensor networks. The main causes of energy loss in a wireless sensor node is the collision of data frames because of simultaneous sending action of different nodes and CSMA/CA overheads, including the additional clear channel assessment. Therefore, two methods to remove additional clear channel assessment and a method to remove collision in wireless sensor networks with star topology and smaller than 256 nodes have been proposed in this thesis which star topology may be a cluster of large network. In methods which have been proposed for additional clear channel assessment , sending of the acknowledgement frame from the coordinator is done only after 192 microseconds so that the rece iver does not clear up more time. In the first proposed method, location of the sending some data frames with a certain length is changed in Backoff. in the second method, one to three zero-byte is added to the end of the data frame according to length of some data frames. To remove the collision , a new algorithm has been proposed as a third method for slotted CSMA/CA , and each node in several specified Backoff performs the clear channel assessment to avoid a collision between data frames. Except for simulatione, implementation and performance verification of the proposed methods, the need to create a sensor network with at least two nodes and one coordinator was essential. Therefore, part of the MAC layer and CSMA/CA algorithm was implemented on FPGA using VHDL hardware language and ARM microcontroller and NRF24L01 transmitter-receiver as the physical layer. After creating wireless sensor network, coordinator LCD shows the sensor amount of nodes temperature and CRC error percentage. In addition to collision percent and reductions in power consumption, parameters of CSMA/CA algorithm can be also changed using the keyboard and nodes are shown on LCD. Keywords CSMA/CA algorithm, Variables BE and CW, Clear Channel Assessment (CCA), Collision of data frames