RMADV-Hop: An Improved DV-Hop Localization Algorithm

With the development of wireless sensor networks (WSNs), a variety of techniques about WSNs receive more and more attention, and the self-localization technique is the prerequisite and basis of these techniques, it also has very important significance on research and practical value. Due to some characteristics and special requirements of WSNs such as limited node energy, the network low-cost requirements and so on, the existing localization techniques are not well suitable for WSNs and the practical application. At present, the researchers have been designed many localization algorithms for WSNs, but these algorithms have certain deficiencies such as low positioning accuracy, high hardware cost etc. Among the various existing localization algorithms, the DV-Hop localization algorithm is easy to implement besides having a large improved space. Analyzing its major shortcomings we proposed an improved DV-Hop localization algorithm—RMADV-Hop (Regular Moving Anchor DV-Hop) localization algorithm.After analyzing the basic idea and localization process of the DV-Hop algorithm we proposed the concept of the regular moving anchor node and the idea of RSSI technology to aiding positioning because of the shortcomings of the DV-Hop algorithm such as the high requirements for the density and the distribution of the anchor nodes, the big error of the per-hop distance, the imprecise estimation of one hop distance, the large cost on the networks hardware, the poor stability of the algorithm and so on.In the RMADV-Hop localization algorithm, the regular moving anchor node can carry out the to and fro movement relying on the fold type route, and finish these tasks during the movement such as the estimation of its self-position, the broadcasting of the localization message, the calculation of the per-hop distance etc. The process of the to and fro movement is that during the first movement, the moving anchor node dynamically connects with the neighbor unknown nodes when staying on the present position in every time or on the border, at the same time it calculates the coordinates of this position according to the movement rout, time and speed, following it records the coordinates and the label of this position, broadcasts the localization message of this position to the neighbor unknown nodes, receives the localization message of the other positions from the neighbor unknown nodes, records and updates the minimal hop count to the other positions, finally, continues to move until arriving at the right border; During the second movement, the moving anchor node calculates the per-hop distance of all positions according to the minimal hop count and distance of each position, then, it returns from the ending position, and broadcasts the message about the per-hop distance of each position by the same way to the first T broadcasting process.The RMADV-Hop localization algorithm solves the following questions using the regular moving anchor node:1. Implementing the dense and even distribution of the localization message. The even distribution of the anchor nodes can not be carried out under the condition of all nodes random distribution, and the number of the anchor nodes is also limited by the network cost. We can safely get the result that the DV-Hop algorithm can not guarantee high positioning accuracy. The positions can be densely and evenly distributed in the network using the to and fro movement of the moving anchor node, and the unknown nodes can receive sufficient localization message in the RMADV-Hop algorithm. So, we could see that RMADV-Hop algorithm improves the average positioning accuracy.2. Reducing the hardware cost of the network. The DV-Hop algorithm needs at least three or even more anchor nodes, and due to each anchor node achieves its self-position by manual layout or the additional hardware, its cost is 100 times than the normal node. So, the density of the anchor nodes increasing can add hardware cost of the network. The RMADV-Hop algorithm only uses one moving anchor node, and the moving anchor node does not need any additional hardware, for it can get its self-position through simple calculation. Furthermore, the RSSI technology has several advantages such as the low cost, easy to realize, without the additional hardware and so on. So, compared to DV-Hop, the RAMDV-Hop algorithm reduces the network cost.3. Improving the algorithm stability. The fold type rout of the moving anchor node can adjust itself to the actual network by adjusting the distance interval D and time interval T. So, the RMADV-Hop algorithm can guarantee the unknown nodes receive sufficient localization message, and keep better localization performance in the sparsely or irregularly distributed network.In the RMADV-Hop algorithm, the function of the RSSI technology to aiding positioning is that, first, assisting the unknown node to judge the nearest position, due to the dense localization message many values of the minimal hop count are 1 between the unknown node and positions, for this situation, the unknown node uses the distance measured by RSSI to judge the nearest position, and uses the per-hop distance of this position as the adjusted value; Secondly, amending the one hop distance value, the two nodes can connect with each other in the range of communication radius according to the connectivity definition, so the distance of such two nodes may be far or near. For this situation, the unknown node compares the adjusted value with RSSI distance, if the correction value is beyond a certain range of RSSI distance, the value of this one hop distance is RSSI distance. This way can reduce the error of one hop distance; further improve the average positioning accuracy.In this paper, we used the OMNeT++ software to do three kinds of the simulation experiment for the DV-Hop and RMADV-Hop localization algorithm. We compared and analyzed the simulation results from the RSSI to aiding positioning, cost, average positioning accuracy and the stability of the algorithm. In the first experiment, we compared and analyzed the function of RSSI technology, the analysis shows that the RSSI technology can further improve the average positioning accuracy. In the second experiment, we analyzed the average position accuracy and hardware cost of these two algorithms, the analysis shows that the RMADV-Hop algorithm can improve the average positioning accuracy of the unknown nodes without increasing the number of the anchor nodes and the hardware cost of the network. The RMADV-Hop algorithm has a little increasing in the communication cost, but it has not increasing in the calculation cost, furthermore it reduces the hardware cost. In the third experiment, we analyzed the stability of these two algorithms, the analysis shows that, the RMADV-Hop algorithm can keep better localization performance in the sparsely or irregularly distributed network.