Note: Methodology for the analysis of Bluetooth gateways in an implemented scatternet

Abstract

This Note introduces a novel methodology to analyze the time performance of Bluetooth gateways in multi-hop networks, known as scatternets. The methodology is focused on distinguishing between the processing time and the time that each communication between nodes takes along an implemented scatternet. This technique is not only valid for Bluetooth networks but also for other wireless networks that offer access to their middleware in order to include beacons in the operation of the nodes. We show in this Note the results of the tests carried out on a Bluetooth scatternet in order to highlight the reliability and effectiveness of the methodology. The results also validate this technique showing convergence in the results when subtracting the time for the beacons from the delay measurements.

Bluetooth Low Power Modes Applied to the Data Transportation Network in Home Automation Systems

“…First, a key point was checked, i.e., the medium range distances considered in the next tests would have no significant influence in the latency of the communication. Then, the methodology to analyze the performance of master-slave gateways in scatternets described in [40] was applied. …”

“…Taking Table 1 as reference, the latency from end-to-end in Bluetooth multi-hop networks when links are always in park low power mode can be modeled as in Equation (1), where some of the time contributions described in [40] are included, namely the amount of hops (N), the inactivity period of the mode (T PARK ), the ratio of inactivity period affecting the latency (α), and the minimum hop latency (δ min ), which summarizes not only the time for processing the data-packets but also other minor contributions. Furthermore, the probability of a retransmission in the multi-hop communication happening is also included in Equation (1): when one retransmission happens ( p 1 (retx) ), two ( p 2 (retx) ), and more.…”

Low Power Multi-Hop Networking Analysis in Intelligent Environments

“…Firstly, it was checked that the medium range distances considered in the following tests had no significant influence in the latency of the communication. After that, it was applied the methodology to analyze the performance of master-slave gateways in scatternets described in [33].…”

“…Taking into account the dependences shown in Figure 6 and Figure 7, the end-to-end latency when T=100ms T=500ms T=1,000ms T=2,000ms T=100ms T=500ms T=1,000ms T=2,000ms gateways nodes are in the sniff mode can be defined as in Equation (2), which considers the time contributions described in [33] and includes the delay time of the first hop (t1sniff), the amount of hops (N), the latency contribution per hop (δa), the inactivity period of the mode (Tsniff), and the ratio of inactivity period affecting the latency (α).…”

“…Taking into account the dependences shown in Figure 4 and Figure 5, the end-to-end latency when gateways nodes are in the park mode can be defined as in Equation (1), which considers the time contributions described in [33] and includes the delay time of the first hop (t1park), the amount of hops (N), the latency contribution per hop (δ1), the latency addition in each hop (2·δ2), the inactivity period of the mode (Tpark), and the ratio of inactivity period affecting the latency (β). Then, the values for the coefficients in Equation (1) were calculated through linear regression to define the curves that best fit to the results of the measurements for the park mode.…”