2015-04-02

It’s generally understood the Internet of Everything eventually will be the interconnect platform of all things, wireless and wireline. The utopian scenario is to have a common platform, with standardized protocols, which everyone builds to. Will that happen? Perhaps, but for a time, especially while the IoE evolves, that won’t be the case. Count on the early IoE being a playing field of disparate protocols that will have to be kluged together via any number of glue technologies and protocols.

On the wireless side, there already are a number of standards-based technologies such as NFC, Bluetooth, RFID and ZigBee, all of which work well, are entrenched, and fairly widely adopted. But there are others, such as medical body area networks, (MBAN), commercial wearables, various vehicle to something (V2V/I/G), WiGig, IBeacon (from Apple), Intelligent Proximity (from Cisco), and others that will be part of the IoE as well. The trick will be for the IoE to seamlessly integrate these technologies across a girth of bandwidths and data rates. This article will look at the major wireless connectivity technologies that are prominent today and where they fit in the big IoE picture.

To license or not to license

On the licensed side, there is much more order and civility. That is because in most cases, licensed spectrum for operators such as cellular, public safety, utility, transportation, intelligent vehicles, radar and satellite is controlled by the government. If you want to play in that space, you have to play by well-established rules. It will not be all that difficult to integrate this segment.

Unlicensed, on the other hand, is pretty much the Wild West of technologies, with no real requirements to play nice. While there are a number of established technologies (as mentioned above), all were developed to operate in a local network, with little thought given to wide-scale interoperability across platforms. These will be the ones that will bring the challenge.

But the first thing engineers need to understand is how most of this works. The remainder of this article will look at the more common technologies, and their metrics.

The four horsemen

The major unlicensed technologies that the IoE will see are Bluetooth, WI-Fi, ZigBee, and 6LoWPAN. Many of the more exotic ones (MBAN, V2x, wearables) are derivatives of these technologies, so they will not be dissected here since what applies to the top level applies to them as well.

Of the four of them, Wi-Fi and Bluetooth are the most prolific and each is morphing new renditions to meet the IoE and next-generation M2M platforms, mainly focused on low power, but the standard versions also will proliferate. (A comprehensive overview of Bluetooth Smart, or low-energy [BLE] can be found here.

Let’s take a look at these four platforms.

1. 6LoWPAN: This is the newest and most interesting because it is particularly applicable to the IoE. It stands for IPv6 over low-power wireless networks, specifically, wireless personal area networks (WPANs). 6LoWPAN refers to the combined protocol stack made up of IEEE 802.15.41 link layer, the TCP/IP stack, and the Internet Engineering Task Force (IETF) IP header. Obviously, IPv6 was the only logical choice and the decision was made not to support IPv4.

This standard was created by a working group within the IETF, specifically RFC 6282, “Compression format for IPv6 datagrams over IEEE 802.15.4-based networks.” It was developed as an adaptation layer between the TCP/IP stack and the link layer and designed so that very low power, limited intelligence (read, processing power) devices can us IP protocols.

As is the case with all such contained networks, devices that will run on 6LoWPAN networks will need an IP-layer gateway, whether it’s wired or wireless, to access the Internet. And, for the time being, that gateway will contain IPv6-to-IPv4 conversion code. The nice thing about 6LoWPAN is that it is being developed with enough foresight to integrate with upcoming innovations that will be part of the IoE, such as mesh networks.

Even so, work is not quite finished. There are some issues that need to be resolved before 6LoWPAN is completely ready for prime time. One of them is that there is, yet, no standard for certification programs for 6LoWPAN solutions. Because there are multiple optional modes in the 802.15.4 link layer, vendors are able to create different types of solutions against the different operational modes under the guise of calling them 6LoWPAN network-compatible. Unfortunately, all the modes are not interoperable at the local network level.

However, there is a built-in ability to communicate over the Internet, regardless of which operational mode is implemented, as long as Internet application protocol used by the various networks is identical. And finally, a 6LoWPAN device can communicate with any other IP-based server or device on the Internet, including Wi-Fi and Ethernet devices.

2. Wi-Fi: This is perhaps the most prolific and best-known wireless technology. Nearly everybody has Wi-Fi. The number of all-inclusive global Wi-Fi hotspots jumped from more than 26 million to nearly 48 million in 2014. That is an increase of roughly 80%, or one new hotspot about every 1.5 seconds in 2014. By 2018 projections are for that number to increase to 340 million hotspots, and most of those will be an access point for the devices of the IoE.

Original Wi-Fi is based on the IEEE 802.11 standard. It was designed, from the ground up, to be a wireless replacement for the IEEE 802.3 Ethernet standard, and specifically for the Internet. It has native integration with TCP/IP for Internet connectivity and will be the predominant wireless interconnect technology in the IoE.

Since the original specification, a number of subsets have been developed that add various functionalities and upgrades. Table 1 shows that progression in major steps. There are also minor developments, such as 802.11 aa-af that deal with various layers, sections, or technologies, and others such as 802.11i that deal with security, for example.


Table 1: Wi-Fi standards to date.

Wi-Fi networks use what is called a star topology (see Figure 1). Early on, each network generally had one access point (AP). Today, that is changing as smart devices now can act as access points as well. It is quite conceivable that a hotspot, such as a coffee shop, medical office, a commercial building, mall or stadium may have any number of APs, all within each other’s coverage area. Fortunately, the Wi-Fi standard includes multiple frequencies so, theoretically, all of the access points should work as long as there are fewer APs than available frequencies.


Fig. 1: Typical Wi-Fi network.

However, it isn’t that simple because there is overlap in the channels within the band, and the distinct bands have different numbers of channels. Plus there is co-channel interference and dynamic AP locations. All of this will be discussed, in depth, in an upcoming article dedicated to Wi-Fi technology.

Most Wi-Fi networks today use complex modulation schemes, typically one from of spread spectrum – frequency hopping (FHSS) or direct sequencing (DSSS). Spread spectrum is well defined a variety of locations in RF design

Wi-Fi has evolved significantly in the last 10 or so years. It now contains encryption (WPA and WEP), requires authentication and the latest technologies and devices are capable of self-configuration and “intelligent” network interoperability.

Exactly how Wi-Fi will fit into the IoE, other than in its present form is still somewhat vague, considering the sheer number of devices that will be using it, and the variants of it under development.

3. Bluetooth: Perhaps the next most common wireless protocol is Bluetooth, which is very popular today in a variety of devices. The standard for this technology is controlled by the Bluetooth SIG. It also has a number of flavors, with the most attention being paid to the next generation dedicated to low power applications, called Bluetooth Smart or Bluetooth Low Energy (BLE) (see related story). The principal technology, however, is the same for both.

Bluetooth originally was developed as a cable replacement for short links, 10 meters or so, operating in the 2.4 GHz ISM band using FHSS. While there are special-use cases that up the ante for capabilities and functionality, 2.0 is the core technology that offers a 2 Mb/s data throughput rate.

A basic Bluetooth network consists of a seven-device, ad-hoc, piconet running in a master-slave configuration. Multiple groups can interface with each other to create a “scatternet” (see Figure 2), where devices in different piconets communicate with other device in other piconets. Note that a master in one piconet may be a slave in another.

In each piconet, the master sets the clock rate and all devices sync to that clock. The basic communication protocol is based on a master clock. The standard timing is 312.5 µs pulse intervals. The simplest configuration is called a single slot. In this case, the master transmits in even slots and receives in odd slots. Each slave receives in even slots and transmits in odd slots. Packets may vary, one, three, or five slots long, but the master always transmits in even slots and the slave in odd.

Today there is a version 3.x, which offers a 24 Mb/s throughput, but it actually uses 802.11 wireless to carry the information. Bluetooth 3.1 only sets up the link. This version offers support for an alternate lower layer. That means that applications that run on earlier versions can be run on this version, but using other technologies such as mentioned above. This is called Bluetooth High Speed (HS) or Enhanced Data Rate (EDR, 3 Mb/s). However, not all 3.0 devices are high-speed or EDR. It depends upon the actual device configuration.


Fig. 2: Bluetooth piconets and a scatternet.

The latest rendition is 4.0 (aka Bluetooth Smart), and is likely to be a key technology for the IoE. The reason is that it supports data collection form very low data rate devices. Its key design parameter is to be able to aggregate a lot of data from a variety of low-periodic sensors such as heart rate monitors, thermostats, appliances, fitness monitors, etc.

4. ZigBee: This has been around since 2003 in its present form, but for a variety of reasons it has taken a while to get traction. First of all, it is similar to Bluetooth in many respects. However, Bluetooth had a much more cohesive development and deployment strategy, and it got most of the attention. Second, ZigBee never had standardization. And third, ZigBee suffered from a lack of interoperability (an open protocol stack and no certification programs).

The initial ZigBee platform was based on the IEEE’s 802.15.4 MAC and PHY layer protocol. On top of that was the ZigBee network layer proprietary protocol stack. The stack is comprised of the transport, application network layers. On top of that sit various sets of domain-specific profiles. These profiles provide device definitions on standard interfaces for product interoperability in different verticals, such as home and building automation, industrial controls, data collection, and routing and alarm systems.

Early ZigBee protocols were software protocols added to the radios rather than being built-in, and they could be used by vendors in very different ways, none of which were interoperable. These early implementations were vendor-specific so, without interoperability, industries where it could work were reluctant to make commitments to the technology.

Another misstep was that ZigBee was mainly positioned as a smart home technology from the beginning. But 10 years ago the smart home was still mostly a concept. It had little implementation and splintered hardware platforms. Pegging ZigBee to the smart home, which had such little momentum until recently, set the development and adoption progress to the same rate as that segment. And once a technology gets positioned, it is hard to move it off the mark. Today that is changing, and as the smart home is gaining popularity ZigBee is coming of age with it.

Moreover, the ZigBee camp has learned its lessons. The ZigBee Alliance has significantly redesigned the ZigBee stack. ZigBee 3.0 has a reworked network and routing layer based ion 6LowPAN and the IPv6 Routing Protocol for Low Power and Lossy Networks (RPL). The reworked product has a single standard that will support interoperability. This is likely to be the turning point for wide-scale adoption of the technology for the IoE, where it has some distinct advantages over Bluetooth, and other short-range technologies (see note 1), and it is designed to support exactly the type of low-power, low-data rate sensor devices that will make up much of the IoE.

Conclusion

Wireless technologies will connect all but the backhaul of the IoE, and some of that as well. The four wireless technologies touched on in this article are going to be the major technologies that will be found within the IoE. No single platform will fit everything. Certain platforms will work better for certain applications that others, but that is an article in itself.

The bottom line here is that for the IoE to interface all things with all other things, these and some other minor wireless technologies will have to interoperate, seamlessly. That is still a slippery slope. It is unlikely that all of these camps will sit down and add interoperability for each other technology natively. The more likely scenario is for them all to operate under open standards that make it convenient to develop gateways and other glue technologies to interconnect them.

That is a reasonable approach. However, caution must be taken to ensure that security and throughput maintained. As we all are keenly aware, more links in the loop, the greater the security risk becomes and the more overhead is required. It will be an interesting ride if you are part of this segment.

Note 1 – Bluetooth, typically has a range of 10m (100m max, with special usage scenarios), while ZigBee’s is 70m. Bluetooth can support a maximum of eight nodes in a point-to-point topology. ZigBee systems can support up to 65,500 devices connected in star, mesh or other topologies, and the various network topologies can be connected to each to form a cluster.

Show more