In the realm of Internet of Things (IoT) systems, accurately forecasting the runtime of Inferencing models on heterogeneous devices is paramount for optimizing resource allocation, particularly in the contexts of Model Distributed Inferencing (MDI) and Data Distributed Inferencing (DDI). This paper delves into the application of Gradient Boosting Regression (GBR) as a predictive modeling technique for estimating runtime in both MDI and DDI scenarios. GBR presents an equitable trade-off between interpretability, robustness against noise, and suitability for moderately sized datasets. The study reviews previous research on IoT inference optimization. It accentuates the multifaceted intricacies of device diversity and the significance of model interpretability within MDI and DDI setups. The primary contribution of this research is the novel application of the GBR model to predict machine learning inferencing runtime in MDI and DDI contexts. This approach is invaluable when empirical data is limited and characterizing the behavior of newly introduced devices is imperative. The paper elaborates on the GBR algorithm’s utilization, hyperparameters, and custom loss functions tailored explicitly for MDI and DDI. The results section exemplifies GBR’s performance across various computational regimes, including MDI and DDI. It offers insights into the model’s balance between accuracy and complexity. A performance comparison with prior models underscores GBR’s efficacy in predicting runtime in MDI and DDI. This work contributes to the ongoing discourse on IoT optimization and predictive modeling.

Over the past several years, there has been a large increase to the number of edge sensing devices gathering various types of data, from images to audio. One of the issues however is getting this data to a device that can quickly perform analysis. Limited network and compute are just a few of the issues that will be taken into consideration, when it comes to offloading data for inferencing. Existing solutions to the limits of edge computation includes, local computation w/ compression and pruned models, cloud offloading, and several distributed methods, including data distributed and model distributed inferencing. Previous work has explored the performance of these options using standard HTTP communication between the edge and external devices, now new network communication protocols are explored specifically gRPC. The impact of utilizing a more lightweight communication protocol to efficiently transfer data from the edge to other devices is explored while specifically focusing on potential drawbacks that may occur. gRPC while shown to perform more efficiently than standard HTTP, however does come at the cost of increased CPU utilization which may impact performance at the edge in select cases. Experiments will be conducted using a testbed which was created to allow definition of both edge, mid-range, and cloud computation devices to be defined each with corresponding compute and memory allocations. Additional, control of the network will be managed via the software defined network allowing for multiple bandwidth cases to be tested for performance comparison in various network constrained environments. Results highlight the performance benefits gRPC in each of the cases showing inferencing speed improvements of the previous HTTP implementation. However several of the cases show better improvements than expected.

Network simulations are widely used for evaluating network functionality and protocols. Large scale network simulations in particular, produce huge amounts of data, making automatic data visualization tools extremely useful. Visual network analysis tools are capable of bringing the salient network performance metrics to the forefront revealing hidden relationships between different entities. This paper describes a custom visualization framework for rapid visual analysis of performance metrics in tactical network simulations. By utilizing network data gathered via the ns-3 simulator, a series of analytical tests using this framework have been conducted. Results from these tests indicate the usefulness of such a visualization framework when studying large scale network topologies.

Tactical wireless networks operate in complex environments with various resource constraints as well as constant link-state changes. These network topologies are predominantly based on some flavor of adhoc network design in order to accomplish rapid deployment goals. Utilizing the ns-3 simulator and a small tactical network topology with 50 nodes we evaluate the performance of both voice and video transmissions. A combination of nix-vector routing and link-state tables are used to create dynamic routing within the simulation. Then the performance of a SIP based real-time audio and video application in an adhoc network is analyzed. Additionally, we have implemented QoS in ns-3 to support traffic prioritization as needed by voice and video applications. Finally we compare the performance of the real-time voice and video applications using different traffic model.

Tactical network environments are complex, resource constrained, highly mobile in nature and performance of the communication links between the nodes is significantly affected by the path loss caused by the buildings. In this paper, we have used the ns-3 discrete event simulator to simulate an urban tactical network environment using the Rosslyn, VA city model. The simulated model was visualized using unity/worldwind. We have calculated path loss data using VPL and injected the path loss data into link selection to visually study the effects of path loss on the link formation in urban environments.