Before starting, I had uh some specific questions. So um do you know that your uh personal information can be useful for a third party? And uh for uh like for example, some companies and some uh applications and do you know uh where and uh how it is started?So this is what will be specifically the talk of today. So um I will cover the use of mobile ed computer and a distributed architecture uh and uh deployed in a mo mobile crowd sensing environment. So uh my name is Hannah Dama. I'm a postdoctoral fellow uh at the Center for Cyber Physical System at Khalifa University in UAE. So um my talks will be generally about introducing the concept of a of things and crowd sensing then um highlighting the difference between centralized and distributed architecture, introducing the concept of each computer and concluding. But by one of our recent research and the recent proposal solution uh that we call off sick. So um as we know that internet of things, it's uh an emerging technology that try to uh connect everyday objects to the internet. So it uh cover all domains of our lives, including health care, home automation, transportation, cybersecurity and industry.

So uh here, as we can distinguish uh that um by the IUT analytics um that is done in 2020 that uh um 1414 projects uh deployed the internet of things uh concepts where uh manufacturing and industry are the most common. One with 22% followed by transportation and then by energy. So he just to look uh to see any of the impact of uh including the internet of things in different domains and how they are uh competing to use this uh this uh this uh emerging technology. So um as we say, in terms of things, it's connecting everyday objects, which means that uh generating a set of a huge number of the data that it will be processed, filtered and uh and stored later. So crowd sensing, it's like a, a specific case or a subdomain of uh uh it's an enabled technology for internet of things that has a main objective is to collect the information or to collect, collect the data uh by deploying a set of users or what we call in the US en crowd.

I think it's workers that use uh mobile devices, generally some smartphones and wear devices and um uh executing some specific tasks while generating or uh collecting um more information about uh a specific uh phenomenon. So there is two kinds or two type of crowd sensing. The first one is the participatory crowd sensing. And the second one is the opportunistic one. So in participatory uh crowd sensing, the users volunteer to participate while this is not the case for opportunistic. Uh The data is censored, collected and shared automatically. And sometimes even without the knowledge of uh uh of the user with participatory, uh the user decide or determine uh how and when and what to sample and where to sample or to collect the data. Why this is not the case for opportunistic that uh gather a large amount of data that will be processed later on. But in a small uh time, it doesn't require too much time. So for participatory um avoid the phone uh issues like for example, uh uh application crash, um bad battery life and um so on and like this kind of issue uh which is not the case for opportunistic. Why? For example, if you um use one of your uh application that you already in uh install in your uh in your uh mobile phone. Uh And when this uh application starts sensing or collecting your information and your um your phone is uh turn it off.

So in this case, the sensing process is stopped because um it is done without the knowledge of the users. And um it is uh based on the the phone conditions also for a participatory, it's uh more expensive or uh require a high cost than the opportunistic one for the same reason because in a participatory as we say that part users, they volunteer to participate while this is not the case for opportunistic uh one.

And those users are sometimes paid for this kind of works for collecting the information. For this reason. It's uh it is more uh expensive, expensive than uh opportunistic crowd sensing. So here, uh this is some uh example of uh mobile applications um like for example, doing on youtube, Facebook, uh Uber, upwork, Instacart and so on that, deploy a set of workers and those workers are paid based on their job. So um here this difference in uh companies uh IUT devices technologies, this is what we call heterogeneous. So a heterogeneous environment is an environment that employ different or diverse um diverse um devices or transmission technologies. For example, there is some devices that can share um for example, we can share our pictures using um for example wifi Direct or Bluetooth or even using Wi Fi or four G. And, and this is a different um completely different transmission technology. And also we can share our information using smartphone as well as using computer or uh for example, dash camera, camera surveillance and so on. So this diversity, this is what we call heterogeneous environment. So in this kind of environment, also the data is heterogeneous. Why?

Because it depends on the devices conditions and the environment conditions and also the transmission technology conditions. So the generated data um after being collected or censored it will be guttered and send it to a centralized platform uh at decentralized platform. This data is uh stored and um requires some additional computation um uh and uh additional processing and uh fil filter filtering. So this uh centralized platform requires some uh or has some challenges or limitations. For example, uh the delay in answering real time requirements because uh as we say, we have a very big uh or uh a very big uh amount of data that is generated a different data, hydrogenous data that is also uh stored. And this kind of um of information need or require more time to be processed and to be ready to send to the end users. So it takes too much time to uh to give a feedback to uh to answer a specific request from, from a specific user. Another challenge is discovering the non connected devices. So um what we mean by non connected devices, it's for example, for smartphone or for other devices, we can use Wi Fi or four G communication network. Uh In this case, uh those devices are accessible by the server and discovered by the server. Like for example, in case of sharing uh using Wi Fi direct or uh Bluetooth uh communication, um those devices can be hidden from the server.

So uh this is one of the main limitation of a centralized platform also um they are complex uh on compute and the data due to, to the the volume of the data that is stored there and uh that is generated. And the last one is the high cost, they are very expensive. So the centralized platform are very expensive because for the same reasons, they require um a strong memory to store the whole information and also require a very high capacity of processing. So um to overcome this uh sum of the limitation of decentralized platform, there is two solutions that was uh recently composed. Um The first one is using a decentralized or a distributed architecture. And this distributed architecture as you can distinguish in these two figures, the first one, the whole information is sent to the cloud. So all the information, whatever the source are sent to a centralized platform. While in the distributed architecture, the information it's distributed according to the area of according to the location of the server and according to the location of the workers or the users itself. So here the information instead of going to a platform, it is distributed in different places or in different services. And this makes the processing of this data very easy and more fast than the case for a cloud or a centralized platform.

So uh the second solution, it's the edge computing. So uh the computing is considered as enabler uh for it services that allow uh computing. Um that is uh an an solution for computing a large amount of data. So it's allow uh it's like an um uh it is used like uh an intermediate layer between the cloud server and the mobile users. So uh it is placed in between. So uh it allow to reduce the complexity of computation because it is placed near to the end users. And also uh it's the, it is responsible to uh do some data filtering, aggregation processing and storage and response um spent in real time to the final users. So uh this computer has a set of benefits. The first one is uh the parallel and the partitioning problem space based on the location as um we said in the previous figures that there is a set of servers and each one of them uh collects uh a set of information that is situated in the same location, um the same location and those information are processed in parallel.

So all the servers work and at the same time uh on the same, not the same information but on different information, but they are uh complementary information. And after that, after this uh computation, if for example, uh the information uh or the request that is sent by the user, um it's a very easy DNA task or it can be answered in a real time. Uh The node is responsible to answer to that user. But in case if the user need more information, this uh the each computer send the request to the cloud computing and the response will be coming back from the cloud computer. But and for small task, the answer or the interaction between the edge server and um the the the mobile user, it's it's it's direct. So uh another benefit uh so this this uh interaction or um uh being placed near to the end user um provide other benefits like reducing or floating the computation for both mobile devices and cloud server, reduce the computational complexity of a centralized cloud and decrease the latency latency and the privacy treat.

So for offloading computation, the age computing as we say, it's placed in intermediate between the mobile devices and cloud server. So uh for example, in a mobile devices, if the data is collected, for example, if I'm using my smartphone and the data is collected uh as we can know that a smartphone has, has a limited DNA uh memory uh storage memory and the capacity of processing. So for example, if I collect a very huge number of information, I cannot process it or compute it in my own uh smartphone. So what I do, I send it to the computer. This can each computer do this uh computation. So in this case, I upload the task that normally should be done in a smartphone, it is already done and uh in each uh computer. And also, for example, if this task um usually it is processed on a cloud computer. But in case of deploying each computer, this ta can be treated locally in each computer without even going to the cloud and getting a feedback from the cloud. So it's like a double benefit of loading the smartphone and also uploading the cloud servers and for the latency, uh this is for sure near to the end users. So it means that there is a fast answer or fast reply to a specific request.

It means that a low latency for the privacy treats. For example, if all the information and this is the case of centralized platform, if all the information is stored on a cloud or a centralized centralized platform, for example, in in case of uh attacks or um for example, material uh destroyed or crashed or something like this, all the information will be lost.

But in case of distributed, the architecture, the information is uh distributed in different places. And for example, if some of this information is missing, um so based on what's still uh other information and the other servers, we can for example, predict what is missing or we can estimate or expect um any uh Domi information. So uh and this is what reduced or increased any uh reduce the privacy threats. So uh here, based uh on uh what we learned about curve sensing about the limitation of centralized platform and about the benefit of each computer and our sensors we try to um to deploy uh or to make like a combination of all these concepts and one architecture. So we proposed a of six solution um which is a computer based um solution that deploy edge nodes, different entities of edge nodes. And it's a distributed architecture and that deploy a set of servers, a set of edge nodes and select a set of workers, which is the third um concept, it's uh deployed in a crowd sensing environment for worker selection or worker recruitment.

So here um I will show you um Yeah. So here uh our solution uh said the following concept first, the research discovery because uh what we, we assume that um our architectures not only deploy the connected devices, the devices that use Wi Fi or uh four G or five G in general wireless communications, but also it can discover other non connected devices, for example, that deploy the Bluetooth.

So because our solution aims to um to uh integrate the concept of heterogeneous um environments. So here we use a different technology also, it's a task allocation because it's based on the task received uh from the server or the worker. Uh Also because we assume also that the task can be initiated by the different entities, maybe a response to a request from each server. Or uh for example, for uneven detection, it will be initiated by the workers. It's a context where because it's based on what we have in the environment, what we have in terms of uh uh IUT devices, uh the connected and the non connected devices also the number of participants or a number of workers available on an area of their location, the location of task and so on.

So all this kind of um uh requirement should be considered uh are considered in our solution also uh heterogeneous worker selection. So um heterogeneous it devices and heterogeneous um uh wireless um or transmission technology lead us to um uh address the concept of heterogeneous workers.

And finally, the budget constraints, why we introduced the concept of budget here because this is the one of the main uh parameters of crowd sensing. So uh crowd sensing, deploy first uh state of workers uh uh mobile devices, as we say, smartphone or we device devices. And um in case of participatory uh uh uh a crowd sensing, we need a specific budget because uh the workers that are recruited and this kind of crowd sensing they are already paid. So it's according to the budget that we assume or that we attribute to each node, uh the selection will be stopped if for example, that budget is reached. So uh we stop the recruitment but um if it is not the case, we keep recruitment for sure. So uh here, this is just uh a video that explains clearly um how the solution works. Sorry. OK. So as we can see here, the final task is the worker selection, it's attributed to each len or local edge node and each cluster. So based on two parameters which are the quality of services provided by each worker. So each len or each uh local not select the best workers based on their quality of services and also based on budgets because uh this is what why we say that our solution is a budget constraint. So uh according to a variable budget, we uh keep recruiting workers.

So as uh uh results uh or a finding, we um we find that our solution selects best workers and uh based on uh that provide a very high quality of services and uh and the quality of services, it include like uh devices characteristic and also workers' reputation. So the workers selected by our solution ha has uh have a high reputation compared to the centralized one and also the information or the data collected from those workers, it's more consistent. So um as a take away uh all what we need to know, it's that mobile crowd sensing, it's a human driven paradigm that um has as a main focus, it's collecting uh the data uh from a physical environment using our mobile devices and a centralized uh mobile crowd sensing. It suffer from a set of limitation that can be overcome by uh the deployment of mobile edge computing. And um without uh forgetting the the concept of heterogeneous environment which is deploying a diverse um entities or uh devices that have a different character and um content.

And if you want to connect and contact me, this is my linkedin my Gmail and Google scholar and copies if you want to read some of my research. And thank you so much.