The evolvable nature of Software Defined Wireless Networking offers great opportunities toward the design and implementation of a Low-cost Network Testbed for Smart Homes. Programmability is an es- sential component on a network gateway to enable efficient management of energy and other network resources for secure, scalable, and cost- effective solutions. In this paper, we studied the programmability of sev- eral SoC and FPGA platforms and introduced a design space for Smart Home Network testbeds by integrating open-source software projects in- cluding OpenWrt, Lede, and OpenFlow to provide intelligent home wire- less networking solutions for Smart Homes scalable to other areas of the Smart Grid and Internet-of-Things technology. We implemented WiFi, BLE, and ZigBee networking on our Low-cost FPGA and SoC platforms and evaluated the network TCP and UDP performance. We conducted a series of experiments on our testbeds and we investigated areas for optimizing this network performance based on novel developments in SDN. Our results provide possible research areas for creating scalable and cost-effective software defined smart home testbeds for advancing Smart Home research.
Smart homes are increasingly becoming available to our daily living. It is pro- jected that over 25 billion IoT devices will be connected by 2020[1]. A Smart home is a communications network linking key electrical appliances and services accessible and monitored remotely. Smart homes can be classified into 2 cate- gories: autonomous houses which act based on sensor conditions, or Intelligent houses which can learn without human intervention. Smart Homes consist of var- ious technologies on the smart grid including wearable technology and a variety of sensors utilizing different protocols. Smart homes can also be centralized or decentralized, however, most systems have adapted a centralized approach where all devices are connected to a single gateway. There are a few challenges in standardizing smart such as energy efficiency, security, and efficient manage- ment. Many solutions have been proposed as seen in chapter 2, however there is another solution worth our attention "Software Defined networking" (SDN) which has started gaining attention in Smart Home research.
SDN is new paradigm shift in networking, it separates the control plane from the data plane, offering network flexibility, introducing programmability and ease of management. SDN is often combined with Network Virtualization to abstract network layers into simple topologies that are task specific. SDN can be incorporated into smart homes, centralizing the network and offering a global network view. SDN can unite different communication technologies, provide better security against external attacks, and offer better energy saving schemes.
In this paper, a Low-cost software defined network testbed for Smart Homes is explored on multiple SoC development boards. The testbed platforms include the Digilent Zybo FPGA with a dual-core ARM Cortex-A9 processor, Intel Galileo Gen 2 with an Intel Quark SoC X1000 application processor, Raspberry Pi 1 Model B+ with the ARM11 CPU, and Raspberry Pi 3 Model B with a quad- core ARM Cortex-A53 CPU (see Table 1). These platforms have much higher processing speeds than conventional routers. We investigated the programmabil- ity of these hardware platforms as a network testbed for Smart Home research to provide connectivity solutions for IoT using SDN.
Software Defined Edge Cloud Network Architecture
The essential components in building a software defined wireless networking testbed for smart home research are shown in Fig. 1. There are four key com-
ponents in a smart home environment, including the network gateway, an SDN
controller, and high bandwidth and low bandwidth communication channels to
manage appliance with efficiency and securely. The data paths to the commu-
nication interfaces between the gateway device and smart home network appli-
ances or external networks are illustrated by grey arrows and wireless links to
the the appliances and home devices are depicted by the bold dashed lines. The
SDN control plane can receive status information from the home appliances and
transmits controls to the appliances through the SmartWLAN and SmartW-
PAN access networks. The controller can also extend security algorithms on all
networks using the OpenFlow protocol to secure the data transmission.
Smart Gateway
To efficiently manage network traffic from an extensible number of connected IoT devices, we integrated “OpenFlow” into our smart home router. OpenFlow is an Open-source Software Defined Networking (SDN) protocol that controls the data forwarding layers of routing [10]. SDN is a systematic shift in the networking architecture where the network control forwarding rule is disabled leaving the data forwarding layers fully programmable. This change in control provides opportunities for applications at the upper layers to seize control of the underlying network allowing them to treat the network as a logical or virtual entity enabling programmability. Quality-of-Service (QoS) network control scripts can be extended onto an SDN controller and improve the multimedia delivery of an embedded OpenFlow Controller.
In this paper we studied the programmability of low-cost SoC platforms with both FPGA non-FPGA architectures as network gateways of a smart home. The gateway device can be implemented on programmable SoC architectures with the latest Linux kernels. The proposed design does not incorporate wired links on the data path to home appliances though wired links are used for linking the gateway to the Internet. Wired communication links are much secure and give a higher throughput, however we consider the modern structure of smart homes and cabling costs of wired systems and integrated the latest wireless commu- nication standards in this design. Wireless systems also have a low complexity during setup and configuration when compared with wired links which is ideal in Home Area Networks (HAN) environments. The proposed smart home gateway accommodates wireless communication networks much more easily by adapting wireless interfaces such as Bluetooth (IEEE 802.15), WiFi (IEEE 802.11) and BLE and ZigBee (IEEE 802.15.4).
Smart WLAN
WiFi has become the most dominant networking technology for implementing wireless local area networks “WLAN”. WiFi technology is built on top of the IEEE 802.11 standard set of media access control (MAC) and physical (PHY) specifications that support applications in computers, smart phones, and other bandwidth sensitive networking devices. WiFi standards include IEEE 802.11a, 802.11b/g/n, and 802.11ac wireless communication standards which operate at the 900 MHz and 2.4, 3.6, 5, and 60 GHz communication frequency bands. “802.11ac” is the latest WiFi standard with dual band support. It supports multiple connections at once and operates at the 2.4 and 5 GHz WiFi frequency bands. 802.11ac is also backward compatible with 802.11b/g/n wireless devices and supports data bandwidth rates up to 1300 Mbps when operating at 5 GHz and 450 Mbps at a 2.4 GHz frequency. Fig. 1 illustrates our proposed software- defined edge-cloud network architecture.
Smart WPAN
BLE and ZigBee are designed for low data rate applications to efficiently con- serve energy. This enables devices and sensors to operate for a number of years depending on the amount of activity and stability of the energy source. These low complexity wireless standards have specifications on both Layer 1 (PHY) and Layer 2 (MAC) and are highly adopted and an anticipated solutions for connecting smart grid devices in IoT. The short range and restricted topology in BLE and ZigBee devices requires mesh and start networking protocols with multi-hop support to overcome the limitations. WiFi radios don’t efficiently man- age power. Using WiFi to manage appliances in the smart home network would exhaust battery powered home appliances much frequently. This has encouraged the development of power efficient home appliances with other wireless technolo- gies. BLE and ZigBee provide a sufficient throughput for smart home network communication in low bandwidth devices, which efficiently utilize energy to im- prove connectivity in Smart Homes.
SDN Controller
As a new method for managing a Smart Home network, we propose Software Defined Wireless Networking (SDWN) to efficiently manage network traffic from an extensible number of connected IoT devices. This is a new research area that has recently gained interest in Enterprise networks, but as far as we know this is the first design integrating OpenFlow on a testbed to conduct experiments for Smart Home research. OpenFlow extends generic features of TCAM switches and routers and provides an open protocol for configuring different Switches and flow tables on Routers. While using OpenFlow to manage the network, researchers can distinguish between experimental streams and workflows to control their own experimental streams by selecting Packets, routing lines, and handling received Packets. This enables experimenting of routing protocols, security models, ad- dress scheduling, and Select IP. The data path of the OpenFlow Switch contains a flow table and an action corresponding to each flow entry. These operations are extensible to a subset of switches for a limited and useful set of operations. The OpenFlow Switch matches each flow entry in the table to a correspond- ing operation with instructions on handling an incoming flow and provides a Secure Channel to link the switch to a remote controller and transmits control commands and packets using the OpenFlow protocol.
In this paper, we presented the design and Implementation of a Smart Home network testbed to conduct parallel experiments on the OpenFlow protocol. The proposed method has been addressed with much detail to encourage more research in Software Defined Wireless Networking and modern communication technologies to support scalable and cost-effective development of Smart Homes, Smart Grid, and Internet-of-Things technology. Our results have proved a very low performance of the software based packet switching module on our testbeds. We admit that much evaluation was not done on the Smart Home testbed due to the current progress. In the future, we wish to give much time to analyzing the performance of our router implementation in a Smart Home scenario for the im- plemented “WiFi”, “BLE”, and ZigBee interfaces. Our experiments will include; energy consumption, network performance, bandwidth utilization, CPU perfor- mance, and other tests to improve all network features for optimal performance. consumption, network performance, bandwidth utilization, CPU performance, and other tests to improve all network features for optimal performance.
This artitle is a summary of the work published in the International Conference on Security, Privacy and Anonymity in Computation, Communication and Storage (SpaCCS 2017) by a group in ITEC-G 2018. It is an extension of project Lion hence the name Lion 3rd Generation. Newer concepts blooming from this project have be published in other articles. More details are available at the end of the article.
