Lecture 17: Wastage Aware Routing

  • Videos
  • May 7, 2021
  • 0
Talha Ibn Aziz – Lecturer at Department of CSE, IUT

Assalamualaikum everyone uh so this is     the 17th     lecture of of the course wireless     networks and the name of today’s topic     is     uh wastage our routing okay so what is     wastage area writing we’re going to see     a little     later on and you’ll see that it’s     mostly or sometimes it’s referred to as     w a r or y okay so what also means     wasted java routing     so let’s start today what we’re going to     do is we’re going to learn about     residual writing from a specific     reference     research paper uh the title of the     research paper is wasted by routing in     energy harvesting wireless sensor     networks okay     so you already know what wireless sensor     networks is     or what ws are but uh you also need to     know what     energy harvesting wireless sensor     networks are okay so we’re going to see     that     on the next few slides so the authors of     this paper are martinez lee and zhao and     the journal where it was published is     ieee sensors journal which is also a     very good journal     and publish and the publication was     available online or i think it was     published on september 2014     so it was about six years ago     so what is wireless sensor network uh     it’s actually network     which has a very uh which has many     challenges and which has very limited     resources     and we’ve already learned a lot about     wireless sensor networks and     some of their some of the mac     protocols of wireless sensor networks in     a previous slide right     and if you remember then one of the most     challenging     issues of wireless sensor networks was     its limited     battery life or its limited lifetime     okay     so uh later on after wireless sensor     networks were     invented or made a specific type of     wireless sensor network     arrows which was energy harvesting     wireless sensor network or ehwsn     so the specialty or characteristics of     this specific     network was that the nodes of this     network could harvest     or collect energy from the environment     so if they were managed properly then it     could potentially give us     a network with infinite lifetime okay     compared to how the nodes could easily     lie in wireless sensor networks now     as as uh harvesting energy is possible     the network can actually exist for an     infinite amount of time     so if you remember there were four basic     or necessary uh let’s say parts of     uh sensor node and there was there were     a few     optional parts as well right so there     was an optional     part which could you know collect power     from the environment okay     so that was the energy harvesting unit     so     the aim in you know building uh energy     harvesting wireless sensor networks is     that it can assure a long time or     a long period of network operation it     can even you know     be infinite okay so there are various     route selection strategies     and as we have more options in     energy harvesting wireless sensor     networks so     we we also have to you know make     protocols which consider both     energy and distance okay so the routing     strategies significantly influence two     things one is     the minimizing the energy consumption of     the     nodes which is just a normal consumption     of the battery life     and also it has to balance the energy     depletion     so that means whenever a node forwards a     packet it uses some part     of its battery life so we also have to     bat uh     balance energy depletion and energy uh     consumption okay so so this means that     uh     if a battery or if a node has very     less battery life then we might have to     send     a package to another path which has     which which might have a longer delay     time but the nodes in that path have     a larger amount of battery life left     okay so     this specific paper is actually a thesis     work which focuses on designing an     energy adaptive routing metric to meet     the     energy harvesting wireless cell center     requirements okay     so the things that we see here you know     minimizing energy consumption and     balancing energy depletion so these     things are already considered in many     other papers     but this paper is special because it     considers something else which is energy     wastage okay     so this is just an introduction of     wireless sensor networks uh     we already know it right a wireless     sensor network is uh     specially distributed autonomous devices     using sensors to monitor physical     or environmental conditions we already     know it and     uh there are many constraints and     limited lifetime is one of them     so the recent advance in energy     harvesting technology gives     promising solution for long lifetime     okay that means that if     the energy harvesting technology is used     properly     the network lifetime can you can     increase by manifolds     so energy harvesting can take many forms     for example it can derive energy from     external sources like     solar energy or thermal energy     wind or kinetic energy or it can uh you     know you also use captured and     stored energy in wireless sensor devices     okay     so usually the solar is the most     easy one i guess and you you might have     already seen a lot of solar powered     uh batteries in in uh your environment     as well     so energy harvesting uh sensor networks     potentially maintain an infinite     lifetime or battery we have already said     that and     solar power wsens collect power every     day if there is sunlight     so this can be used uh very easily but     the thing is     at night time there is no sunlight right     so     yeah you have to consider uh this factor     and you have to     collect as much energy as possible so     so even though it might seem like the     energy is unlimited because the sun is     always there     for a short time of or for short term     period     the energy is actually not infinite     because when there is no more sun     that means maybe at night it can only     perform     a specific amount or limited amount of     transmissions     after that the energy will be depleted     similarly even during daytime if the     amount of transmissions done by a node     exceeds the amount of energy it can     harvest at a time     then also the node node may die out okay     so in long term the energy may be     unlimited but in short term the energy     is limited     and and also uh the nature of the power     sources is also stochastic or     random because uh you don’t know how     much energy     the node may collect at a given time     because even though     it might be daytime it might be a cloudy     day it might be a rainy day and so on     so the energy distribution is also     uneven     because each node may have their own     harvesting rate or might     be in a place where they receive less or     more     energy so the research objective of     energy harvesting wireless sensor     network is     uh to ensure that the network goes on     for an infinitely long amount of time     okay     so how to do that we have to use the     harvested energy appropriately so that     it is feasible to meet the goal and the     goal is     a time network operation and     to ensure the network operation for a     long time     maximization of network remaining energy     is important     so this means after the any operation     is done uh it must be ensured that     the maximum amount of energy is left in     the nodes     so we have to select that route or uh     the specific action which ensures that     after the action is done the overall     network     energy level is the highest okay and     another thing is     maximization of minimum energy levels in     network     so the first one was total network     energy and this one is the     minimum energy level so that means let’s     say if i pass     a packet through a specific route maybe     the overall energy used is very less     but a specific node in that route     has a very less amount of battery life     left     so it might be a better option to     actually use another route where     all the nodes have a significant amount     of     energy left okay so these are two things     which we have to     consider and these are the two uh     promising     factors which researchers usually look     into     so uh energy harvesting wireless sensor     networks has many applications     and as you can now you know restore your     energy     you can also process or you can transmit     a higher volume     of data compared to before and     in these networks as well the data is     periodically forwarded to sync     and examples can be precision     agriculture wildlife monitoring we have     already seen these examples     and the requirement is as we have     already said     for you know perpetual network operation     the requirement is     you have to maximize the total network     remaining energy and the minimum energy     levels as well okay     so what is routing metric arriving     metric is unit calculated by a routing     protocol for selecting or rejecting     that means for selecting or not     selecting a routing path     for transferring data traffic and     examples we have already seen the     previous lecture hop count     expected transmission count right so for     uh for satisfaction     of certain energy objective of uh energy     harvesting wireless sensor network     the routing metric should be aimed at     these two things that     the things we have already discussed     before right so in this scenario the     routing metric is     based on energy levels instead of the     delay     so there are already various uh existing     route selection schemes in energy     harvesting wireless sensor networks     and they consider a lot of things uh     they can they might consider the     battery residual energy so this means     the     uh amount of energy left in the node     after a transmission is done     then they also uh you know optimize the     predicted     harvest energy that means they try to     predict the amount of     energy that a node harvests in a given     time     and following or considering that it     performs the     let’s say uh transmissions or selects     the routes     and it might also consider the estimated     energy consumption     so that means it estimates the amount of     energy that is used     when a transmission is done and not     the amount of energy left okay so you     might confuse the first and third point     but they’re actually quite different and     then also the channel condition might be     considered     so all these things they are already     done before     this paper was proposed and they are all     very significant     uh you know objectives but uh when this     paper was written     there was no uh significant     research done on overcharge of battery     and over charge of battery is not very     unusual     so this paper specifically targets over     charge of battery or we can also call it     wastage of battery     so to understand wastage of battery we     have to see a few examples     so we can say that although the amount     of energy we get from     sun can be considered or for any ambient     uh an environment can may be considered     infinite     but the battery capacity of the node is     finite okay     so whenever ah an excess amount of     charge     is added to a finite capacity battery     then we get     a wastage of energy so let’s see for     example this     is a battery and we can say or we can     assume that the     green part of this battery is uh the     energy level     of the battery and the white part is the     unfilled     or the part which does not have any     battery left     so if the total capacity of the battery     is one     and if the current energy level is 0.9     then we can say that 0.1 is the amount     of     battery energy which has been used     already so delta t here     is the prediction horizon so this means     the amount of energy it     can gain after a specific amount of time     so in this scenario we can say that the     harvested energy is     0.2 so once uh     you add 0.2 and 0.9     what do you get you actually get 1.1     but the total capacity of the battery is     one so this means that the extra part     0.1 battery     life is wasted okay so this is the     concept of     battery wastage now uh considering     the wastage of battery life isn’t can be     taken as an research objective as well     and it can you know provide us with an     optimal solution as well according to     this paper     and this can be done by maximizing     the utilization of harvested energy so     this means that we have to send data     packets in such a way     so that this wastage is minimized and     not just     the overall battery usage or let’s say     uh the residual battery left not only     that     but this paper also considers the     wastage and it tries to minimize the     wastage     so let’s see how the previous case can     be avoided     if the battery is used so we again go     back to the previous scenario where the     battery     size is one and the present energy is     0.9     and we can predict that it will uh you     know harvest     delty amount of time so delta amount of     uh energy so the harvested energy is 0.2     and obviously the wastage was supposed     to be 0.1     but now if we actually consume due to     transmission     that part that means 0.1 battery life     then     there is no wastage okay so that means     if we     you know uh use this node for     transmission instead of any other node     then its wastage can be minimized     okay so this is what this paper aims to     do     so how does this actually change the     route okay     so to understand the effect of this     protocol on     uh routing we have to see a few another     example which will be shown later on so     to understand that example we have to go     through these things     the first one is there are two     components of wastage our     routing uh protocols okay so the metric     is     uh the metric can be considered     according to energy consumption     due to packet transmission okay so this     means the amount of energy that is     used when a packet is transmitted     okay and the second thing is the total     network energy wastage due to battery     over charge     so this uh paper considers these two     things     one is the usual energy consumption and     the second thing is also the energy     wastage due to battery     over charge so how does it solve or how     does it consider these two things it     chooses a route from a set of routes     which ensures a minimum of total network     energy wastage     plus the path energy consumption due to     transmission so it adds adds the two     factors that we just mentioned     it adds them up and tries to minimize     their sum okay     so the total network energy wastage due     to selection     of let’s say route omega i     is the on path wastage of omega i     plus the off path wastage of omega i     so omega i is the route through which     the     packet is uh transmitted and on pathways     charge means     the wastage along the path     through which the packet has passed and     off path wastage means     the wastage in all the other parts which     are     not part of this route okay so we’re     going to see the example so     no need to get confused so let’s say in     this scenario uh the source is     s and the destination is d so the route     is in this uh case one and two so the     so the data packet passes from s through     one and two and then which is d or     destination     so the wastage on path for route     for this specific route can be     considered as the wastage     which is which happens in node 1 and     node 2 after utilizing those nodes okay     so that means when the data packet is     passed through these nodes     the wastage that is done in node one and     two is considered as wastage on path     and if we have another route which is     not used okay     so in this scenario the other route is     three four five but that is not being     used     the transmission is going through node     one and two     so if there is any wastage on those     nodes for example in three four and five     then that is called the wastage off path     okay     so so you can you should understand this     by now right so if the path if the route     was going to three four five then the     wastage on path would be     the wastage of node one of node three     four and five     and the way stage off path would be uh     for node     one and two is it clear so what i’m     saying is just the opposite of this     scenario     okay so the total network energy wastage     due to the selection of the route uh     omega 1 which is the route 1     2 is the on path wastage added with the     off path wastage okay so this is the     routing metric that     this paper uses okay     so let’s calculate the cost for this     example     so uh the route cost for     path sigma is path transmission cost     plus on path wastage in all the used     node     and off pathway stage on all the     unused node okay so path transmission     cost means the amount of energy used     for transmission on pathway stage in all     use node means     the amount of energy wasted in all the     nodes in the     path and off path wastage on all unused     node means     the wastage in all the nodes which are     not in the path okay     so in the first node the battery level     is 0.8 and after harvesting the battery     level has increased by     0.1 so uh     the wastage if this node is not used     is zero okay and if it is used     for transmission then each transmission     uses up     0.1 battery life for this node so if     it it is used for transmission then it     will lose     0.1 battery life which means     even after uh transmission and even     before transmission     there is no wastage okay so ew on one     means     if this node is used then the wastage is     zero     and ew of one means if this node is not     used     then the wastage is 0 so the battery     life     for node 2 is 0.9 and similarly the     energy harvested here is     0.1 okay so we can understand from this     as well that even if this node is used     or not used in both cases the wastage     should be     zero because it also uses 0.1 battery     life to send a transmission okay     so for both node 1 and 2 even if both     the nodes are used or not used     uh the wastage is zero okay so     if you calculate the route cost here the     route cost for path transmission is 0.1     and     0.1 they they should be added right     because uh let me just mark it here     because this is the path uh a link     transmission cost for this link and this     is the transmission cost for     this link so if we sum them up then this     is what we get this is the     path transmission cost okay     then let’s go through the other nodes so     for node 3     the energy level is 0.9     and the harvest rate is 0.2     so now we can understand right     previously uh     it had only 0.1 battery life left     which can be filled but after harvesting     it is harvesting 0.2 battery life so     that means 0.1     will be wasted if this node is not     being used and if it is being used     then that wasted energy will be used so     then there will be     no wastage right you can see here so the     off path wastage is 0.1 that means if     this node is not used     then the wastage for this node is 0.1     and     if this node is used the transmission uh     cost is 0.1 so if this is used then     there will be no wastage     so then for node 4 the battery life is     0.85     and the harvest weight is 0.2 so     similarly     if uh this node is not used that means     the off pathway stage is     0.05 and if it is     used then it uses up 0.1 battery life     for a transmission     so after using 0.1 battery life there is     no more wastage right     so the on path wastage is zero and     for this node the battery life is 0.6     so it harvests about 0.2 battery life     and if it is not used then there is no     wastage because after harvesting the     energy level reaches     0.8 and     the transmission cost for this node is     0.1     so even after transmission there is no     energy     wastage okay so now if we calculate for     all these nodes     then we get two calculations for the     first one     that means if when the     packet is going through this path     then if you can calculate then the path     transmission cost     for this part is point one here zero     point one here okay so we get zero point     one and     zero point one and on path wastage that     means the wastage for     node one and two is zero right we can     see the on path here is zero     and on path here is also zero and     the off path wastage here is 0.15 why     because there is some wastage here 0.1     for off path and 0.05     here for off path so if you add them up     then the wastage is 0.15     and if you add all of these up then the     route cost is     0.35 now let’s calculate     if we send the data through this path     then what do we get     so this path is uh three four five or     sigma three four five     so the path transmission cost is     one here one here and one here     so that means zero point three and the     on pathway stage     is zero because uh this uses up the     energy     wasted energy here and the wastage     energy here as well     so the on path wastage is zero zero and     zero     so there is no on path wastage here but     the off path wasted is also zero here     because even if you do not use the nodes     one and two here     even then there are there is no wastage     here     so as a result we can see that the     overall route cost for this route is     0.3 so we can understand from this     example that     even though the path transmission cost     for     this path is only 0.2 and this one is     0.3     that means according to the path     transmission cost this is the best path     but if you consider both wastage and     transmission cost then this is the     better path because this gives us the     value 0.3     okay so please uh go through this     example again     and if you have any questions you can     ask me okay     so we can clearly say that the path     sigma 345     that means this path leads or gives us     the minimum cost and maximum     remaining energy okay     so this should be our best path and not     one two     so that was the end of this lecture so     if you have any questions     feel free to ask in all the platforms     you already know about the platforms     right     so that’s it assalamu alaikum everyone          


Lecture 16: Routing Metrics

Lecture 18: Routing Protocols Part 1

Leave a Reply

%d bloggers like this: