Trust and Performance Data

Cloudlet computing is one of the applications of edge computing and sometimes cloudlet nodes are unable to meet users’ resource requirements so in such scenarios, instead of shifting the request instantly to remote cloud, concept of cloudlet federation is introduced. Cloudlet federation is a collaborative framework for sharing resources, scalability, and load balancing between the different members of a cloudlet. The research [1] proposed a cloudlet federation model based on performance. However, this conventional cloudlet federation model lacks the parameter of trust which is very vital to avoid exploitation by malicious nodes. This research [2] aims to incorporate trust in a conventional cloudlet federation model and evaluate both performance and trust of cloudlet nodes. Trust model in this research composed of three metrics that include availability, reliability and success rate. Experiments are performed to measure the performance and trust of both conventional cloudlet federation model and the trust-aware cloudlet federation model. The provided dataset is a spreadsheet titled Trust and Performance Data.xlsx that contains around 40,000 instances for the calculation of each metric. This dataset compiles comprehensive performance metrics each contributing to a holistic understanding of network performance. Along with network performance, this dataset also contains information related to trust metrics to ensure the credibility of the cloudlet nodes in a federated environment. Cloudlet A, C and D are available in LAN environment while Cloudlet B and E are in the MAN environment. The considered performance metrics are latency, hop count, bandwidth, offload time, throughput and transfer time. Latency: Latency measures the time taken for data to travel from the source to the cloudlet nodes. It is computed with the help of ping time and measured in milliseconds. Hop Count: Hop count denotes the number of intermediary devices (routers or switches) that data packets traverse during transmission. Lower hop counts typically signify potentially faster data delivery. Bandwidth: Bandwidth refers to the maximum rate of data transfer over a network connection. Offload Time: Offload time measures the duration required to delegate/transfer tasks or data processing to specific cloudlets. Throughput: Throughput represents the actual data transfer rate achieved in a network, considering factors like file size and offload time. Transfer Time: Transfer time calculates the total time taken to transmit a specific amount of data from a source to a cloudlet node. It is calculated with the help of latency and offload time and it is measured in seconds. The considered trust metrics are availability, success rate and reliability. Availability: Accepted Jobs / Total Jobs where total jobs are the tasks a cloudlet need receives while accepted jobs are the ones that the cloudlet node accepts by considering its resources. Success rate: Successful Jobs / Accepted Jobs where successful jobs are the tasks that cloudlet node successfully completed while accepted jobs are the ones that the cloudlet node accepts by considering its resources. Reliability: It is a cumulative factor that is calculated with the help of CPU utilization, memory utilization, hardware utilization and bandwidth utilization. The percentage values of these utilizations are converted to data values between 0 to 1. w1, w2 and w3 are different weights assigned to the trust parameters for the calculation of Cumulative Trust Value (CTV). If you are using this dataset, please cite the following papers. Citation: Nayyer, M. Z., Raza, I., & Hussain, S. A. (2020). Cfro: Cloudlet federation for resource optimization. IEEE Access, 8, 106234-106246. Paper Link Shahid, V., Nayyer, M. Z., & Ahmed, U. (2023, March). Trust-aware Cloudlet Federation Model for Secure Service Selection. In 2023 International Conference on Energy, Power, Environment, Control, and Computing (ICEPECC) (pp. 1-5). IEEE. Paper Link Contribution: The conventional cloudlet federation model is proposed by Dr. Muhammad Ziad Nayyer. This research adds the parameters of Trust in this proposed model to ensure the trustworthiness of a cloudlet node and to avoid malicious access. References: [1] Nayyer, M. Z., Raza, I., & Hussain, S. A. (2020). Cfro: Cloudlet federation for resource optimization. IEEE Access, 8, 106234-106246. Paper Link [2] Shahid, V., Nayyer, M. Z., & Ahmed, U. (2023, March). Trust-aware Cloudlet Federation Model for Secure Service Selection. In 2023 International Conference on Energy, Power, Environment, Control, and Computing (ICEPECC) (pp. 1-5). IEEE. Paper Link