Heuristic Algorithms for Co-scheduling of Edge Analytics and Routes for UAV Fleet Missions

Aakash Khochare,Yogesh Simmhan,Francesco Betti Sorbelli,Sajal K. Das

DOI: https://doi.org/10.48550/arXiv.2102.08768

2021-02-07

Abstract:Unmanned Aerial Vehicles (UAVs) or drones are increasingly used for urban applications like traffic monitoring and construction surveys. Autonomous navigation allows drones to visit waypoints and accomplish activities as part of their mission. A common activity is to hover and observe a location using on-board cameras. Advances in Deep Neural Networks (DNNs) allow such videos to be analyzed for automated decision making. UAVs also host edge computing capability for on-board inferencing by such DNNs. To this end, for a fleet of drones, we propose a novel Mission Scheduling Problem (MSP) that co-schedules the flight routes to visit and record video at waypoints, and their subsequent on-board edge analytics. The proposed schedule maximizes the utility from the activities while meeting activity deadlines as well as energy and computing constraints. We first prove that MSP is NP-hard and then optimally solve it by formulating a mixed integer linear programming (MILP) problem. Next, we design two efficient heuristic algorithms, JSC and VRC, that provide fast sub-optimal solutions. Evaluation of these three schedulers using real drone traces demonstrate utility-runtime trade-offs under diverse workloads.

Robotics,Distributed, Parallel, and Cluster Computing

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to plan the missions of a drone fleet in the city to complete the activities regularly provided by users, while maximizing the total utility and satisfying the energy and computing constraints of each drone. Specifically, the paper proposes a novel Mission Scheduling Problem (MSP), aiming to coordinate the flight paths of drones between multiple waypoints and on - board edge - computing tasks, enabling drones to complete as many activities as possible with a single full charge, while ensuring that these activities are completed within the specified time and maintaining efficient use of energy and computing resources.

### Problem Background
With the development of drone technology, unmanned aerial vehicles (UAVs) or drones are increasingly widely used in smart cities, such as in traffic monitoring, building surveys, and package delivery. Autonomous navigation allows drones to visit waypoints and complete specific tasks, and common tasks include hovering and observing a location using an on - board camera. Advances in deep neural networks (DNNs) enable these videos to be analyzed for automatic decision - making. Drones also have on - board edge - computing capabilities and can perform DNN inferences during flight.

### Problem Definition
The problem proposed in the paper is the Mission Scheduling Problem (MSP) of the drone fleet, with the goal of maximizing the utility obtained from activities under the premise of meeting activity deadlines, energy, and computing constraints. Specifically, MSP needs to solve the following key issues:
1. **Path Planning**: Determine the optimal path for drones to visit each waypoint.
2. **Task Scheduling**: Reasonably arrange on - board analysis tasks of video data on drones.
3. **Energy Management**: Reasonably allocate the battery capacity of drones among flight, hovering, and computing tasks.

### Problem Complexity
The paper proves that MSP is an NP - hard problem. This means that it is infeasible to find the optimal solution in polynomial time, so efficient heuristic algorithms need to be designed to find approximate solutions.

### Solutions
To address the complexity of MSP, the paper proposes the following methods:
1. **Mixed - Integer Linear Programming (MILP)**: By modeling MSP as a mixed - integer linear programming problem, the optimal solution can be found, but this method is only applicable to small - scale inputs.
2. **Heuristic Algorithms**: The paper designs two efficient heuristic algorithms, namely JSC (Job Scheduling Centric) and VRC (Vehicle Routing Centric), which can quickly provide sub - optimal solutions and are suitable for large - scale inputs.

### Evaluation
The paper evaluates these three algorithms using actual drone trajectory data, showing the trade - off between utility and running time under different workloads.

### Main Contributions
1. **Definition of System and Application Models**: Describes in detail the drone system model, application model, and utility model.
2. **Formal Definition of MSP**: Proposes a formal definition of the mission scheduling problem.
3. **Proof of NP - Hardness**: Proves that MSP is an NP - hard problem.
4. **MILP Optimal Solution**: Designs an optimal solution algorithm based on MILP, which is suitable for small - scale inputs.
5. **Heuristic Algorithms**: Designs two efficient heuristic algorithms, which are suitable for large - scale inputs.
6. **Performance Evaluation**: Through the evaluation of actual drone trajectory data, shows the performance of the algorithms.

In summary, by proposing MSP and designing corresponding solutions, this paper provides a theoretical and practical basis for efficient mission planning of drone fleets in the city.