Estimation of Control Variable for Obtaining Discreet Time-variant Delays in Grid Networks Heerok Banerjee* 1 , S Murugaanandam 2 , V Ganapathy 3 Student 1 , Assistant Professor 2 , Professor 3 Department of Information Technology, SRM Institute of Science & Technology, India heerokbanerjee_ra@srmuniv.edu.in 1 , murugaanandam.s@ktr.srmuniv.ac.in 2 , ganapathy.v@ktr.srmuniv.ac.in 3 Abstract— In several data-intensive networking applications, Autonomous Decentralized Systems (ADS) are observed to achieve higher efficiency and reliability. ADS are defined as hard real-time systems comprising of several compound modules that operate independently to achieve a common target of the system. We have previously proposed a decentralized networking architecture entitled “Autonomous Collaborative Decentralized-Community Network (ACDCN)” which primarily targets to optimize high-latency optimal relay networks by adopting a collaborative computing paradigm. The proposed architecture is a hybridized model consisting of distributed and decentralized networking mechanisms operating on large-scale data . In this paper, we have illustrated the parameter estimation methodology to estimate the collaborative index ‘σ’ that can allow moderating the rate of collaboration in the proposed model. In general, we have addressed a more specific challenge viz. to attain a certain degree of synchronous heterogeneity in homogeneous real-time systems. The simulation of the model was performed using OMNet++ with varying collaborative indices and network parameters. The results were plotted using QtiPlot and the data was visualized using maltab. Keywords- Autonomous Decentralized Systems, Distributed Computing, Grid Computing, Non- Parametric Estimation I. INTRODUCTION In 2016, the annual global IP Traffic was reported to be 1.2 ZettaBytes per year. As per the global IP Traffic forecast, the global IP Traffic is expected to reach 3.3 ZB by 2021 [see 1]. Also, The traffic generated from wireless and mobile devices accounts more than 63.2% of traffic and India contributes 13.22% percent to the global IP traffic. CISCO predicts that the overall IP traffic will grow with a cummilative annual growth rate (CAGR) of 24% from 2018 to 2022. These figures implicate that our digital needs have exceedingly impacted the infrastructure of the internet into yielding autonomous networked systems with multi-objective capabilities, which are primarily targetted to serve massive bulk requests and perform huge computations parallely. But such systems do have hardware limitations and can only perform upto a threshold quantity. The exponential increase in internet usage and mobile data, along with the growing user demand for digital connectivity has introduced numerous challenges for academicians and researchers. After decades of research, many architectural shifts and technological innovations have resulted in yielding stable networking architectures and has provided solutions to numerous networking challenges. But these solutions have, however, faced formidable deployment barriers due to high configuration demands and thus, have led researchers to focus on multi-objective and cost-effective networking schemes that balances the trade-off between connectivity and performance. Also, the increased internet traffic was observed to impact the QoS and the performance of such autonomous networked systems like Content Delivery Networks (CDNs), Software Defined Networks (SDN), etc. So, in the current state-of-the-art, there are several major issues that needs to be addressed particularly in the wake of designing and implementing large-scale data-intensive networking architectures. In this paper, we have discussed about Autnomous Decentralized System (ADS) based Grid networks, which is a multi-objective, resource-aware, reliability based networking architecture targetted primarily to compute on the mass of the data provided by the end user. The internet has always been considered as a single entity, by the majority of its subscribers. But its constituents have never been accounted towards the remodification of the forthcomming trends in networking and microservice architectures, namely CDNs and SDNs. Several micro- service architectures have been proposed, but they have been found to be inconsistent in delivering a balanced service model particularly in the wake of networking security and cost-effective performance. The fundamental resources that are shared in such data-intensive networking applications are typically bandwidth, buffers and processing memory. In this paper, we have determined that even with a fair allocation of shared resources, it is insufficient to satisfy the network requirements when the topology is centralized. We have discussed an novel mechanism to substitute effective scheduling in ADS-based networking architectures by introducing a control parameter, collaborative index, denoted by ‘σ’ and have given a blueprint on estimating the range and domain of this variable by using non-parametric estimation technqiues. We have discussed about optmial and saturated collaborative states, which introduces alternative approaches than scheduling in distributed systems and autonomous networks. We have also simulated the mechanisms involved in ADS systems and have analysed the performance and security-measures of the ADS-Grid based networks with varying time-delays and homogeneous configuration of component hosts.