الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
Advances in communication, electronics, and digital technologies have led to creation of low cost, low energy consumption, multi-functions, and tiny size Sensor Node (SN).
These SNs can remotely sense natural phenomena (as: heat and humidity), can manipulate data, and can communicate, which led to idea of Wireless Sensor Networks (WSNs), which supported a lot of valuable applications (e.g. in commerce, health, and military fields).
Advances in WSN resulted in large number of protocols regulating energy consumption, specially: clustering protocols, which differ accqrding to the network architecture, and require application.
This research suggests two S-table Heterogeneous S-ensors Network protocols: SHSN-l and SHSN-2 for modifying homogenous Stable Sensor Network (SSN) protocol. The two new proposed protocols are applied within two-level Heterogeneous WSNs (HWSNs). The first proposed protocol SHSN-l suggests new weighted probability equation for electing Cluster Heads (CHs) which depends on initial SN energy. The second proposed protocol SHSN-2 suggests new weighted probability equation for electing CHs based on the ratio between the residual SN energy in on one hand and the average network energy on the other hand.
The two newly suggested protocols were simulated using a special developed C++ program. The simulation was performed for the proposed protocols on one hand, and Stable Election Protocol (SEP), Deterministic Cluster-Head Selection (DCHS), and SSN protocols on the other hand under same environment conditions. Results of the simulation were compared. SHSN-l showed about 25%, and 16% improvement in First Node Dead (FND), and Half Nodes Dead (HND), respectively, than the original SSN, while SHSN-2 showed about 33% and 21% improvement in FND, and HND, respectively, than the original SSN along same metrics. In addition, both of the proposed protocols showed generally some improvement in the network stability in the rest of the other protocols (SEP, and DCHS). Moreover, both proposed protocols increased the average throughput during the stability period (FND and/or HND) compared to current clustering protocols.