Serverless Edge Computing is an extension of the serverless computing paradigm that enables the deployment and execution of modular software functions on resource-constrained edge devices. However, it poses several challenges due to the edge network's dynamic nature and serverless applications' latency constraints. In this work, we introduce EdgeFn, a lightweight distributed data store for the serverless edge computing system. While serverless comput- ing platforms simplify the development and automated management of software functions, running serverless applications reliably on resource-constrained edge devices poses multiple challenges. These challenges include a lack of flexibility, minimum control over management policies, high data shipping, and cold start latencies. EdgeFn addresses these challenges by providing distributed data storage for serverless applications and allows users to define custom policies that affect the life cycle of serverless functions and their objects. First, we study the challenges of existing serverless systems to adapt to the edge environment. Sec- ond, we propose a distributed data store on top of a Distributed Hash Table (DHT) based Peer-to-Peer (P2P) Overlay, which achieves data locality by co-locating the function and its data. Third, we implement programmable callbacks for storage operations which users can leverage to define custom policies for their applications. We also define some use cases that can be built using the callbacks. Finally, we evaluate EdgeFn scalability and performance using industry-generated trace workload and real-world edge applications. / Master of Science / Serverless Edge Computing is an extension of the serverless computing paradigm that enables the deployment and execution of modular software functions on resource-constrained edge devices. However, it poses several challenges due to the edge network's dynamic nature and serverless applications' latency constraints. In this work, we introduce EdgeFn, a lightweight distributed data store for the serverless edge computing system. While serverless comput- ing platforms simplify the development and automated management of software functions, running serverless applications reliably on resource-constrained edge devices poses multiple challenges. These challenges include a lack of flexibility, minimum control over management policies, high data shipping, and cold start latencies. EdgeFn addresses these challenges by providing distributed data storage for serverless applications and allows users to define custom policies that affect the life cycle of serverless functions and their objects. First, we study the challenges of existing serverless systems to adapt to the edge environment. Sec- ond, we propose a distributed data store on top of a Distributed Hash Table (DHT) based Peer-to-Peer (P2P) Overlay, which achieves data locality by co-locating the function and its data. Third, we implement programmable callbacks for storage operations which users can leverage to define custom policies for their applications. We also define some use cases that can be built using the callbacks. Finally, we evaluate EdgeFn scalability and performance using industry-generated trace workload and real-world edge applications.

This thesis presents an innovative strategy for distributed Graph Neural Network (GNN) training, leveraging a peer-to-peer network of heterogeneous edge devices interconnected through a Distributed Hash Table (DHT). As GNNs become increasingly vital in analyzing graph-structured data across various domains, they pose unique challenges in computational demands and privacy preservation, particularly when deployed for training on edge devices like smartphones. To address these challenges, our study introduces the Adaptive Load- Balanced Partitioning (ALBP) technique in the GraphDHT system. This approach optimizes the division of graph datasets among edge devices, tailoring partitions to the computational capabilities of each device. By doing so, ALBP ensures efficient resource utilization across the network, significantly improving upon traditional participant selection strategies that often overlook the potential of lower-performance devices. Our methodology's core is weighted graph partitioning and model aggregation in GNNs, based on partition ratios, improving training efficiency and resource use. ALBP promotes inclusive device participation in training, overcoming computational limits and privacy concerns in large-scale graph data processing. Utilizing a DHT-based system enhances privacy in the peer-to-peer setup. The GraphDHT system, tested across various datasets and GNN architectures, shows ALBP's effectiveness in distributed GNN training and its broad applicability in different domains and structures. This contributes to applied machine learning, especially in optimizing distributed learning on edge devices. / Master of Science / Graph Neural Networks (GNNs) are a type of machine learning model that focuses on analyzing data structured like a network, such as social media connections or biological systems. These models can help identify patterns and make predictions in various tasks, but training them on large-scale datasets can require significant computing power and careful handling of sensitive data. This research proposes a new method for training GNNs on small devices, like smartphones, by dividing the data into smaller pieces and using a peer-to-peer (p2p) network for communication between devices. This approach allows the devices to work together and learn from the data while keeping sensitive information private. The main contributions of this research are threefold: (1) examining existing ways to divide network data and how they can be used for training GNNs on small devices, (2) improving the training process by creating a localized, decentralized network of devices that can communicate and learn together, and (3) testing the method on different types of datasets and GNN models, showing that it works well across a variety of situations. To sum up, this research offers a novel way to train GNNs on small devices, allowing for more efficient learning and better protection of sensitive information.

In recent years, Internet-of-Things (IoT) devices generate a large amount of personal data. However, due to the privacy concern, collecting the private data in cloud centers for training Machine Learning (ML) models becomes unrealistic. To address this problem, Federated Learning (FL) is proposed. Yet, central bottleneck has become a severe concern since the central node in traditional FL is responsible for the communication and aggregation of mil- lions of edge devices. In this paper, we propose Scalable Federated Learning via Distributed Hash Table Based Overlays for network (Scaled) to conduct multiple concurrently running FL-based applications over edge networks. Specifically, Scaled adopts a fully decentral- ized multiple-master and multiple-slave architecture by exploiting Distributed Hash Table (DHT) based overlay networks. Moreover, Scaled improves the scalability and adaptability by involving all edge nodes in training, aggregating, and forwarding. Overall, we make the following contributions in the paper. First, we investigate the existing FL frameworks and discuss their drawbacks. Second, we improve the existing FL frameworks from centralized master-slave architecture by using DHT-based Peer-to-Peer (P2P) overlay networks. Third, we implement the subscription-based application-level hierarchical forest for FL training. Finally, we demonstrate Scaled's scalability and adaptability over large scale experiments. / Master of Science / In recent years, Internet-of-Things (IoT) devices generate a large amount of personal data. However, due to privacy concerns, collecting the private data in central servers for training Machine Learning (ML) models becomes unrealistic. To address this problem, Federated Learning (FL) is proposed. In traditional ML, data from edge devices (i.e. phones) should be collected to the central server to start model training. In FL, training results, instead of the data, are collected to perform training. The benefit of FL is that private data can never be leaked during the training. However, there is a major problem in traditional FL: a single point of failure. When power to a central server goes down or the central server is disconnected from the system, it will lose all the data. To address this problem, Scaled: Scalable Federated Learning via Distributed Hash Table Based Overlays is proposed. Instead of having one powerful main server, Scaled launches many different servers to distribute the workload. Moreover, since Scaled is able to build and manage multiple trees at the same time, it allows multi-model training.

Grid computing systems are suffering from reliability and scalability problems caused by their reliance on centralised middleware. In this thesis, we argue that peer to peer middleware could help alleviate these problems. We show that peer to peer techniques can be used to provide reliable storage systems, which can be used as the basis for peer to peer grid middleware. We examine and develop new methods of providing reliable peer to peer storage, giving a new algorithm for this purpose, and assessing its performance through a combination of analysis and simulation. We then give an architecture for a peer to peer grid information system based on this work. Performance evaluation of this information system shows that it improves scalability when compared to the original centralised system, and that it withstands the failure of participant nodes without a significant reduction in quality of service. New contributions include dynamic replication, a new method for maintaining reliable storage in a Distributed Hash Table, which we show allows for the creation of more reliable, higher performance systems with lower bandwidth usage than current techniques. A new analysis of the reliability of distributed storage systems is also presented, which shows for the first time that replica placement has a significant effect on reliability. A simulation of the performance of distributed storage systems provides for the first time a quantitative performance comparison between different placement patterns. Finally, we show how these reliable storage techniques can be applied to grid computing systems, giving a new architecture for a peer to peer grid information service for the SAM-Grid system. We present a thorough performance evaluation of a prototype implementation of this architecture. Many of these contributions have been published at peer reviewed conferences.

We present a distributed data structure, which we call "RootChord". To our knowledge, this is the first distributed hash table which is able to adapt to changes in the size of the network and answer lookup queries within a guaranteed two hops while maintaining a routing table of size Theta(sqrt(N)). We provide pseudocode and analysis for all aspects of the protocol including routing, joining, maintaining, and departing the network. In addition we discuss the practical implementation issues of parallelization, data replication, remote procedure calls, dead node discovery, and network convergence.

We present a distributed data structure, which we call "RootChord". To our knowledge, this is the first distributed hash table which is able to adapt to changes in the size of the network and answer lookup queries within a guaranteed two hops while maintaining a routing table of size Theta(sqrt(N)). We provide pseudocode and analysis for all aspects of the protocol including routing, joining, maintaining, and departing the network. In addition we discuss the practical implementation issues of parallelization, data replication, remote procedure calls, dead node discovery, and network convergence.

This work presents a novel Peer to Peer (P2P) framework for Session Initiation Protocol (SIP) on Mobile Ad Hoc Network (MANET). SIP is a client-server model of computing which can introduce a single point of failure problem. P2P SIP addresses this problem by using a distributed implementation based on a P2P paradigm. However, both the traditional SIP and P2P SIP architectures are not suitable for MANETs because they are initially designed for infrastructured networks whose most nodes are static. We focus on distributed P2P resource lookup mechanisms for SIP which can tolerate failures resulting from the node mobility. Our target application is SIP-based multimedia communication in a rapidly deployable disaster emergency network. To achieve our goal, we provide four contributions as follows. The first contribution is a novel P2P lookup architecture based on a concept of P2P overlay network called a Structured Mesh Overlay Network (SMON). This overlay network enables P2P applications to perform fast resource lookups in the MANET environment. SMON utilizes a cross layer design based on the Distributed Hashing Table (DHT) and has direct access to OLSR routing information. Its cross layer design allows optimizing the overlay network performance during the change of network topology. The second contribution is a distributed SIP architecture on MANET providing SIP user location discovery in a P2P manner which tolerates single-point and multiple-point of failures. Our approach extends the traditional SIP user location discovery by utilizing DHT in SMON to distribute SIP object identifiers over SMON. It offers a constant time on SIP user discovery which results in a fast call setup time between two MANET users. From simulation and experiment results, we find that SIPMON provides the lowest call setup delay when compared to the existing broadcast-based approaches. The third contribution is an extended SIPMON supporting several participating MANETs connected to Internet. This extension (SIPMON+) provides seamless mobility support allowing a SIP user to roam from an ad hoc network to an infrastructured network such as Internet without interrupting an ongoing session. We propose a novel OLSR Overlay Network (OON), a single overlay network containing MANET nodes and some nodes on the Internet. These nodes can communicate using the same OLSR routing protocol. Therefore, SIPMON can be automatically extended without modifying SIPMON internal operations. Through our test-bed experiments, we prove that SIPMON+ has better performance in terms of call setup delay and handoff delay than MANET for Network Mobility (MANEMO). The fourth contribution is a proof-of-concept and a prototype of P2P multimedia communication based on SIPMON+ for post disaster recovery missions. We evaluate our prototype and MANEMO-based approaches through experimentation in real disaster situations (Vehicle to Infrastructure scenarios). We found that our prototype outperforms MANEMO-based approaches in terms of call setup delay, packet loss, and deployment time.

An introduction to Chord Algorithm.

Source is converted into pdf format. An introduction to Chord Algorithm.

Consistency preservation of replicated data is a critical aspect for distributed databaseswhich are strongly consistent. Further, in fail-recovery model each process also needs todeal with the management of stable storage and amnesia [1]. CATS is a key/value datastore which combines the Distributed Hash Table (DHT) like scalability and selforganization and also provides atomic consistency of the replicated items. However beingan in memory data store with consistency and partition tolerance (CP), it suffers frompermanent unavailability in the event of majority failure. The goals of this thesis were twofold (i) to implement disk persistent storage in CATS,which would allow the records and state of the nodes to be persisted on disk and (ii) todesign nodes failure recovery-algorithm for CATS which enable the system to run with theassumption of a Fail Recovery model without violating consistency. For disk persistent storage two existing key/value databases LevelDB [2] and BerkleyDB[3] are used. LevelDB is an implementation of log structured merged trees [4] where asBerkleyDB is an implementation of log structured B+ trees [5]. Both have been used as anunderlying local storage for nodes and throughput and latency of the system with each isdiscussed. A technique to improve the performance by allowing concurrent operations onthe nodes is also discussed. The nodes failure-recovery algorithm is designed with a goalto allow the nodes to crash and then recover without violating consistency and also toreinstate availability once the majority of nodes recover. The recovery algorithm is based onpersisting the state variables of Paxos [6] acceptor and proposer and consistent groupmemberships. For fault-tolerance and recovery, processes also need to copy records from the replicationgroup. This becomes problematic when the number of records and the amount of data ishuge. For this problem a technique for transferring key/value records in bulk is alsodescribed, and its effect on the latency and throughput of the system is discussed.