In this paper, we propose ERGO (edge architecture for Ag-IoT), an edge-computing architecture for infrastructureless smart agriculture environments. We also develop Ag-IoT application APIs and the associated microservice infrastructure. Our implementation and evaluations show that ERGO can operate independently of cloud-backed assistance, is highly scalable, modular, and affords composability benefits to Ag-IoT systems. We also demonstrate that ERGO outperforms traditional infrastructure in response latencies and transactional throughput, on average, by over 54% and 77%, respectively.

The Interplanetary Network (IPN) or the Interplanetary Internet is a network composed of interconnected space objects, which are in turn connected to mission control stations on the surface of Earth. The IPN is our only portal to the deep space, and yet it has been relatively sparse, until recently. With the ongoing and the planned missions to the outer space, the Delay Tolerant Networking (DTN) based network infrastructure will require more scalable routing and scheduling algorithms. In this paper, we propose the first Mixed Integer Linear Programming (MILP) model for message routing and scheduling in the IPN using Multi-Attribute Decision Making (MADM) principles. Based on this model, we propose a novel MADM-based algorithm called Multi-Attribute Routing and Scheduling (MARS) algorithm. This algorithm uses a sliding window of size n to schedule the first n messages in the buffer based on multiple attributes. After finding the optimal schedule for these messages (in terms of delivery rate), they are routed using our proposed Dijkstra-based routing algorithm. We use an existing MADM technique, PROMETHEE II, and consider the four main attributes of a message: size, priority, time to live (TTL), and time in buffer (TiB). Finally, we run multiple simulation experiments in order to test the performance of the proposed MARS and show that MADM coupled with scheduling and routing in IPN delivers at least three times more messages than a previously proposed technique, the Contact Graph Routing (CGR), while significantly reducing the average end-to-end delay and overhead.

Humans' deep space exploration missions have attracted great interest in the past decade. The Interplanetary Network (IPN), hence, has to undergo many structural changes, as the traffic loads and types will drastically increase. In this paper, we propose N-Look Ahead Routing and Scheduling (N-LARS), a novel scheduling and routing technique for DTN-based IPNs with these new challenges. Our N-LARS model for Delay Tolerant Networking (DTN)-based IPNs provides scheduling of bundles (also called requests or messages) and efficient routing across the overall network. We develop a mixed integer linear programming (MILP) formulation for N-LARS problem and solve it using CPLEX. We further derive a heuristic that scales better than the MILP for large networks. We run experiments on networks mimicking the proposed Mars project by the European Space Agency. Finally, we compare the performance of the network using our scheduling model to its performance using the usual routing technique such as the Contact Graph Routing (CGR). Our simulation results show that the overall performance of the network using N-LARS for bundle scheduling and routing is four times better than CGR, in terms of delivered bundles, and 13.8% closer to the sub- optimal schedule generated by the CPLEX solver.

Interplanetary Networks (IPN) are classified among challenged networks and hence are hard to model using static graphs. Furthermore, they do not behave optimally when operated using the standards and techniques of static networks. Delay Tolerant Networking (DTN) is one of the suggested solutions to overcome these networks' challenges. The more widely used implementation of DTN uses Contact Graph Routing (CGR) to find the path from source to destination. In this paper, we identify the shortcoming of CGR that results from overlooking the future contacts and propose the Earliest Arrival Optimal Delivery Ratio (EAODR) Routing that examines all the paths both with the desired earliest departure time and in the future in order to choose the earliest arrival path from a given node. EAODR finds the route that delivers the exchanged message (a.k.a. bundle) at most at the same time as CGR's route. We propose a Modified Temporal Graph (MTG) model that provides a near-real-time representation of the deterministic dynamic networks. We base EAODR routing algorithm on the MTG model. Our results show that we can reduce the delay by 12.9% compared to CGR when we apply our algorithm to over 50 combinations of bundle sizes and transmission times.

Interplanetary Internet or Interplanetary Networking is envisaged as a space network which interconnects spacecrafts, satellites, rovers and orbiters of different planets and comets for efficient exchange of scientific data such as telemetry and images. In this paper, we implement a layout of the Interplanetary Internet (IPN) with the Interplanetary Overlay Network (ION) software module that uses Contact Graph Routing (CGR). The experiments are then implemented on the Global Environment for Network Innovations (GENI) testbed. Along with realistic contact plans (CP) of the nodes, this network implementation was used to run experiments testing the performance of Delay Tolerant Networking (DTN) with and without cross links between Mars orbiters. The experiments showed that in an Earth-Mars communication network using two Mars orbiters, allowing cross links between the orbiters results in increasing the amount of data transferred by roughly 9.2%. Data sent from Mars Rover to the Earth stations also increases by 35.7% when a third satellite (Mars Express) was added to the network without cross links. Finally, when cross links are allowed across all satellites orbiting Mars and serving as relay nodes between the Earth stations and Mars rover, the communication was enhanced by almost 46%. We conclude that by adding cross links, the performance of the network is enhanced for a better transmission of data from Mars to the Earth, which is very pertinent for the scalability of the network.

Fixed-point format is essential to most efficient Digital Signal Processing (DSP) implementations. The conversion of an algorithm specification to fixed-point precision targets the minimization of the implementation cost while guaranteeing a minimal processing accuracy. However, measuring such processing accuracy can be extremely time consuming and lead to long design cycles. In this paper, we study reference approaches to measure fixed-point errors of Linear Time-Invariant (LTI) systems without feedback. Unsurprisingly, we find the existing analytical approach significantly faster than a straightforward simulation-based estimation. However, we also show that such analytical approach can incur high estimation errors for some particular bitwidth configurations. Accordingly, we propose a new hybrid approach, which is able to reduce by up to 4 times the error of the analytical estimation, while still being more than 10 times faster than the simulation-based estimation.