Event-Driven Bandwidth Allocation with Formal Guarantees for Camera Networks

Abstract

Modern computing systems are often formed by multiple components that interact with each other through the use of shared resources (e.g., CPU, network bandwidth, storage). In this paper, we consider a representative scenario of one such system in the context of an Internet of Things application. The system consists of a network of self-adaptive cameras that share a communication channel, transmitting streams of frames to a central node. The cameras can modify a quality parameter to adapt the amount of information encoded and to affect their bandwidth requirements and usage. A critical design choice for such a system is scheduling channel access, i.e., how to determine the amount of channel capacity that should be used by each of the cameras at any point in time. Two main issues have to be considered for the choice of a bandwidth allocation scheme: (i) camera adaptation and network access scheduling may interfere with one another, (ii) bandwidth distribution should be triggered only when necessary, to limit additional overhead. This paper proposes the first formally verified event-triggered adaptation scheme for bandwidth allocation, designed to minimize additional overhead in the network. Desired properties of the system are verified using model checking. The paper also describes experimental results obtained with an implementation of the scheme.

Model Checking a Self-Adaptive Camera Network with Physical Disturbances

“…This means that the network manager chooses each element b i such that n i=1 b i = 1 and tries to maximize the quality of the camera streams. For the resource allocation, we use the algorithm proposed in [26] adapted to network bandwidth in [33], because it guarantees properties like starvation avoidance and fairness among the different streams.…”

“…In fact, the encoded frame size, using a movement-based encoder (like MPEG), depends on many different factors, including (among others) nature, lighting conditions, and the amount of movement detected and encoded in the scene [16]. Even when images are transmitted as simple JPEG frames, the encoded scene makes a difference in the image sizes [33]. This motivates the need to include some considerations about the physics in the network bandwidth distribution protocols.…”

“…Transmission protocols, like the Flexible Time Triggered for Switched Ethernet (FTT-SE) [30], guarantee the non-violation of the assigned bandwidth. In a recent paper [33], we designed and implemented a scheme for bandwidth allocation and camera adaptation that decouples the two inherent adaptation dimensions: (i) bandwidth distribution, and (ii) adaptation of encoding parameters. The encoding parameters are used to adjust the frame sizes and trade the quality of the resulting stream for its compression.…”

“…It is well known that multiple control policies can negatively impact the performance of the system [21], so we needed formal guarantees that this would not happen in this case. We therefore resorted to model checking, and verified properties like the transmission of frames [33]. A model checker ensures that the bandwidth allocation protocol allows cameras to transmit their frames (of a given size).…”

“…• We incorporated disturbances in both a probabilistic and a stochastic model, expanding on [33]. • We compared the probabilistic and the stochastic model exposing their tradeoffs (capturing real-life behaviour versus the computational complexity of the verification process).…”