One idea is to compute or estimate energy consumption in your network with the first protocol. Then, employ the second protocol and compute energy consumption again. In general, nodes in WSN have three main states: listening, transmitting and sleeping. These values may be obtained even from nodes datasheets. If you determine the time spent by  your nodes in each state, you can compute the total energy consumption for each node, and then for the whole network. I recommend the attached publication. I hope it helps.

Conference Paper Measuring Energy Consumption in Wireless Sensor Networks Using GSP

One idea is to compute or estimate energy consumption in your network with the first protocol. Then, employ the second protocol and compute energy consumption again. In general, nodes in WSN have three main states: listening, transmitting and sleeping. These values may be obtained even from nodes datasheets. If you determine the time spent by  your nodes in each state, you can compute the total energy consumption for each node, and then for the whole network. I recommend the attached publication. I hope it helps.

Conference Paper Measuring Energy Consumption in Wireless Sensor Networks Using GSP

Another alternative is to simulate the protocols and the scenarios of their operations and then compute the energy consumptions of each protocol. Gradually change the scenarios and compute the energy consumptions again and again.  One can then implement the protocols using motes and then record the energy consumption. One can then compare results obtained through simulations and practically implementing them using motes. To make a solid piece of work, try to analyze their behaviour mathematically and try to provide reasons for their variations.

Best way to showcase your work on energy conservation  is to simulate your work on any comfortable event driven simulators againt QoS or Network performance evaluation parameters like PDR, throught, end-end-delay, time etc.  then conduct the testbudd on different plotforms, check out the results and plot the graph using GNUplot/Xgraph/MSexcell 07 etc and  claim your work. Another thing need to do parally is to define a precise mathamatical model for your energy related work and called it as a energy model prove the result on matlab or scilab...

The main power consumption sources of any WSN are the transimission and the reception of the data. therfore your main idea should be to reduce the sent and recieved data. in addition to reduce the number of nodes that are active and ready to send and recieve if you can gurantee that in any protocol then you can increase the WSN lifetime and reduce the power consumption

In addition to above suggestions, mainly  energy is used for Computations, Sensing, and transmitting, energy for each section needs to calculated and tweeked by applying standard algorithms and compare with your algorithms. Then analyse what is best !

being interested, i'm looking for research topic in the WSN. could anyone guide me how to start researching about it such as which protocol is interesting or which simulator. thank you in advance.

Energy for Routing Path, Computation, Environment Sensing etc, OMNET++/NS2 can be used to simulate these works.

Thanks very much for all of you

LEACH and HEED protocols

Your question on improvement on energy consumption specifically depends upon which class of protocols you are taking about. For WSN, Preamble Sampling is one of the efficient and low power class of MAC protocols in which duty cycling concept is used. Generally these are asynchronous protocols and energy consumption not only depends upon transmit, liten, receive, transitions states but also on collisions and re-transmission. If you are more interested in this kind of evaluation and improvements, attached is an interesting hardware/software based model.  

Node in WSN mainly spend energy on transmit, receive, overhear , on internal operations, retransmission, and listen for connection. Draw a precise mathematical model for all these energy consumption parameters and analyse the model using some statistical inputs and verify the output. Secondly develop your own scheme and analys the result with recent work and draw the conclusion based on some network parameters like PDR, Delay, jitter, remaining energy, and energy consumed etc.. .  if your scheme outperforms existing standard scheme then you can proven your work.

based on clustering of nodes in same route

Energy is fixed, The way you schedule your work will minize your energy consumption

Clustering is also another way but if you go for Dominat set of nodes you can utilize maximum work with minimum energy efficiency

The best way is to implement both schemes on an experimental testbed and then compare the total energy consumption, or, if the nodes are battery-powered, let both protocols run until the batteries are depleted and compare the lifetime. See, e.g., the last experiment reported in

M. Haenggi and D. Puccinelli, “Routing in Ad Hoc Networks: A Case for Long Hops,” IEEE Communications Magazine, vol. 43, pp. 93-101, Oct. 2005.

In other sections, this paper uses a combined analytical/experimental approach to demonstrate the energy benefits of routing over long hops. You may find some of the techniques useful in your own work.

Thank you all so much

Yes, I agree with Maria. However, you can keep it simple and just compute the amount of power consumed on a node by taking into consideration the traffic load and distance traveled of each packet. Keep deducting the power consumed to receive and send messages.  Once the simulation is over just deduct power consumed due to communication from the original amount on that node. In the end add all the remaining power left on each node to get the total network remaining power. The reason why you can do this is because communication costs are one of the biggest contributors to power consumption.

The transmission of the power consumed in transmitting and receiving a k-bit message, a distance d (using this radio model), can be  computed as:

E_Tx (k,d)=E_elec× k + ε_(amp )×k×(d)^2

  and

E_Rx (k)=E_elec× k 

Hello Moahmmed. One key issue related to energy consumption in WSN is the topology of the network: how many active nodes and active links the network has at a given moment. Controlling and reducing the topology to an acceptable level of service by keeping nodes in deep-sleep mode or by reducing the transmission range are feasible solutions to this problem, plus it keeps resources to extend the lifetime of the network. Check on topology construction and topology maintenance and you'll find many techniques to do this.

Best,

PW

First draw the state model. Energy is used in several states:

- Transmitting

- Receiving

- Sensing

- Idle (Ready)

-Sleep

- Overheads ( Biggest area for improvement): Its a broad area and contains lots of parameters and metrics. Most papers mention this topic.

Energy consumption are usually compared against the following metrics:

- packet arrival rate (data generation interval, or inter-arrival time)

- number of nodes in the network ( vary the #)

- packet size

- range ( and hops in case of multi-hop network)

A simple method of development is as follows:

1. Develop the proposed model (design and mathematical analysis )

2. Simulate the proposed model

3. Compare the results of both analysis and simulation

4. Choose two or three state of the art available protocols and adjust the parameters according your model.

5. Compare the results obtained from those protocols with your models.

6. The graphs will show the differences.

An example paper can be found here: Mikko Kohvakka, Jukka Suhonen*, TimoD Hämäläinen and Marko Hännikäinen, Energy-Efficient Reservation-Based Medium Access Control Protocol for Wireless Sensor Networks, EURASIP Journal on Wireless Communications and Networking 2010, 2010:878412 doi:10.1155/2010/878412

A easier way to prove the improvement in energy consumption is to show the increase in network life time. First simulate a known WSN scenario using ns2 and record the network lifetime. Then implement the protocol in ns2 and using the same scenario record the lifetime of the network i.e after how much time the nodes die. This when shown in a graph  will show the improvement in energy consumption. This is the way generally WSN protocols are shown to be superior in energy consumption over other WSN protocolsn in Journal papers. 

Conference Paper Improvement of Energy Consumption and Load Balance for LEACH...

http://www.ijcsmc.com/docs/papers/January2014/V3I1201405.pdf

https://arxiv.org/ftp/arxiv/papers/1202/1202.1892.pdf

https://www.hindawi.com/journals/js/2016/9278701/

Conference Paper Improved Leach Communication Protocol for WSN

http://ieeexplore.ieee.org/document/7861171/