Wireless sensor network (WSN) typically has energy consumption restriction. Designing energy-aware routing protocol can significantly reduce energy consumption in WSNs. Energy-aware routing protocols can be classified into two categories, energy savers and energy balancers. Energy saving protocols are used to minimize the overall energy consumed by a WSN, while energy balancing protocols attempt to efficiently distribute the consumption of energy throughout the network. In general terms, energy saving protocols are not necessarily good at balancing energy consumption and energy balancing protocols are not always good at reducing energy consumption. In this paper, we propose an energy-aware routing protocol (ERP) for query-based applications in WSNs, which offers a good trade-off between traditional energy balancing and energy saving objectives and supports a soft real time packet delivery. This is achieved by means of fuzzy sets and learning automata techniques along with zonal broadcasting to decrease total energy consumption.
A wireless sensor network (WSN) is composed of a large number of sensor nodes, which are densely deployed either inside the phenomenon or very close to it. Since the sensor nodes are often inaccessible, the lifetime of a sensor network depends on the lifetime of the power resources of the nodes. Power is also a scarce resource due to the size limitations [
Many routing algorithms and protocols have been proposed for different types of WSNs (i.e., [
Rumor is an energy saving protocol that provides an efficient mechanism combining push and pull strategies to obtain the desired information from the network [
This paper presents a routing strategy applicable to various forms of query-based applications that offers a reasonable trade-off between the energy saving and balancing. More precisely, we propose an energy-aware routing protocol for query-based applications in WSNs called the ERP, which is supported by learning automata and fuzzy sets and which uses zonal broadcasting to decrease the total energy consumed. Our initial results demonstrate the potential and the effectiveness of ERP, making it a promising candidate for a number of WSN applications.
The rest of the paper is organized as follows. In Section
More specifically, ERP satisfies the following design objectives.
ERP is a query-based routing protocol designed to consider both energy and distance while routing packets across a network. It balances the load among the different sensors with a twofold goal: preventing the sensors from running out of battery while keeping the routes to reach the destinations relatively short. ERP can be divided into two phases: query broadcasting and data forwarding (Figure
EPR routing mechanism: (a) query broadcasting and (b) data forwarding.
When a WSN starts its operation, station
After receiving the query packet, each intermediate node will save information of the query sender (i.e., energy level and also sender and the station position) into its Neighbor List. The components of the Neighbor List for each neighbor are as follows:
After inserting information of the query packet into the Neighbor List, the intermediate node checks the DP field. In first time, the DP field of the received query packet is null and the intermediate node only inserts its ID, position, and energy level and rebroadcasts the query packet. When destination
In the next rounds of query sending, for finding destination
If we consider that destination
The radius is computed based on the following:
In the equation, Velocity is the average speed.
Using the ED, the ERP protocol will compute a limited query zone to reduce energy consumption of broadcasting.
We propose a new query zone with an optimum size (see Figure
Proposed query zone.
The nodes receiving the query packet can forward or discard it depending on their location. For instance, in Figure
If
When the destination
Procedure of selecting next hop in data forwarding phase is based on learning automata (LA) and fuzzy set (FS) techniques. LA and FS offer their idea about selecting next hop individually and finally one decision maker (DM) algorithm makes final decision.
As we mentioned before, Neighbor List includes M1 and M2 fields. M1 field is filled by learning automata (LA) and M2 field by fuzzy set (FS). Each node that receives a data packet runs its DM algorithm. The DM looks at the Neighbor List and finds the neighbor with highest probability in the M1 field (called
The processing of the DM module is summarized in Algorithm
(1) (2) If ( (3) send the packet to the selected neighbor; (4) // If IDs do not match then run tie-breaking rules; (5) (6) choose (7) (8) choose (9)
Method of filling M1 and M2 fields by LA and FS is described in the following.
The theory of fuzzy sets was introduced by Zadeh in 1965 [
The ERP algorithm considers a fuzzy set
Membership function computation: the membership function consists of three factors,
The membership degree of the
When a node (i.e., node
The rationale of using expression (
The basics of the mechanism are illustrated in Figure
Updated probability scheme.
Phase 1
Phase 2
Thus, in fact, the probability updating will be enabled if the LA and the DM select the same neighbor as the next hop. When node
In our model, we considered four behavioral cases for rewarding or penalizing a neighbor
In the first case, the energy-distance-hop relationship is below the average, and thus the learning automata in
In this section, we evaluate the ERP’s performance by comparing it to the following routing protocols: Rumor [
We used a surface that was 1000 m × 1000 m. The radio range was set to 177 m, with an available bandwidth of 2 Mbps and a radio transmission (TX) power of 4 dBm. Each simulation had a 4-hour duration, and the tests were run under various conditions, such as with different amount of sensors, namely, 1000, 1200, and 1400 nodes, and also with 10 different amounts of seeds. Moreover, the placement of the sensors in the terrain and their initial energy levels were selected randomly. It is worth highlighting that even though the placement and initial energy of the nodes were set randomly, once set, those factors remained fixed for rest of the trials to obtain comparable results across experiments. In the simulations presented here, the traffic in the network is always initiated by a source station
In this scenario, we assume a critical situation, where the energy levels for transmission mode are very low. Under these conditions, we evaluate the different routing schemes considering three different tests.
In this scenario, the energy levels of the sensors are set sufficiently high so as to avoid experiencing node failures during the simulation runtime. Our goal in this case is to compare the fairness in terms of energy consumption. In order to avoid bias in the comparison, we ensure that all the routing schemes transmit the same amount of data and that this occurs without node failures. We carry out three tests to examine how the routing schemes save and manage energy in regular operation mode.
The most commonly used measure of network lifetime is the time until the first node runs out of battery energy. In ERP, the first sensor fails after ~8900 to ~10500 seconds depending on the number of nodes present in the network (see Figure
Figure
In this scenario we evaluate the quality of these protocols in normal energy situations. Figure
Actually, beyond merely comparing the particular values obtained in each figure, the most important conclusions that can be extracted from the tests as a whole are basically the following. The results show that the combination of a FS technique and LA can improve energy balancing and, more importantly, that the combined operation (ERP’s use of fuzzy set plus learning) can work better than using only one technique (EQR).
In this paper we studied energy-aware query-based routing protocols. From the routing perspective, we have observed that the current destination-initiated query-based routing protocols can be considerably improved, especially, if we aim for a better balance between the energy savings and energy balancing objectives. We have proposed a new energy saver/balancer routing protocol. We simulated and compared our routing protocol with traditional Rumor and newer EQR protocols. Indeed, four different types of tests were carried out and described, and in most of these tests it was indicated that the ERP obtained significantly better results.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This work was supported by the EU ITEA 2 Project: 11012 “ICARE: Innovative Cloud Architecture for Real Entertainment.”