%0 Journal Article 
%A Cruz de la Torre, Carlos
%A Palomar González, Esther
%A Bravo Muñoz, Ignacio
%A Gardel Vicente, Alfredo
%T Cooperative demand response framework for a smart community targeting renewables: testbed implementation and performance evaluation
%D 2020
%@ 1996-1073
%U http://hdl.handle.net/10017/43232
%X Demand response (DR) is emerging as the workhorse of achieving energy efficiency
and reducing our carbon footprint, which persists as a major challenge amongst all the different
energy-chain players, i.e., the utility providers, policy makers, consumers, and the technology sector. For instance, the Internet-of-Things (IoT) paradigm and network-enabled appliances/devices have
escalated the expectations of what technology could do for the acceptance of DR programs. In this
work, we design, deploy on a scalable pilot testbed, and evaluate a collaboration-based approach
to the demand-side management of a community of electricity consumers that jointly targets green
consumption. The design of the framework architecture is centralized via the so-called aggregator,
which optimizes the demand scheduled by consumers along with their time frame preferences
towards the maximization of the consumption of renewables. On the pilot, we opt for lightweight,
yet efficient platforms such as Raspberry Pi boards,and evaluate them over a series of network
protocols, i.e., MQTT-TLS and CoAP-DTLS, paying special attention to the security and privacy of the
communications over Z-Wave, ZigBee, andWiFi. The experiments conducted are configured using
two active Living Labs datasets from which we extract three community scenarios that vary according
to the flexibility or rigidity of the appliances’ operation time frame demand. During the performance
evaluation, processing and communication overheads lie within feasible ranges, i.e., the aggregator
requires less than 2 s to schedule a small consumer community with four appliances, whereas the
latency of its link to households’ controllers adds less than 100 ms. In addition, we demonstrate
that our implementations running over WiFi links and UDP sockets on Raspberry Pi 4 boards are
fast, though insecure. By contrast, secure CoAP (with DTLS) offers data encryption, automatic key
management, and integrity protection, as well as authentication with acceptable overheads.
%K Cooperative demand response
%K Consumption scheduling
%K Renewable supply
%K Raspberry Pi board
%K Performance evaluation
%K CoAP
%K MQTT
%K TLS/DTLS
%K Electrónica
%K Electronics
%~ Biblioteca Universidad de Alcala