Wireless 2025: A look at wireless in the year 2025
By Kevin Fitchard
A decade and a half ago, the transition from analog to digital cellular created an explosion in wireless services. What started as an expensive service for a privileged few blossomed into a global industry, which today counts 4 billion subscribers worldwide. Wireless connections overtook wireline ones. For many people in developing countries, the mobile device became their first phone and their first means of accessing the Internet.
Today we’re witnessing a new revolution in which wireless has come to signify data services as much as voice. Over the next two years, the first 4G networks will emerge, pairing mobility with true broadband for the first time. The handset has begun to evolve from a mere phone into a miniature multimedia computer, and portable and mobile devices with no voice capabilities to speak of have started connecting to the wide-area cellular network.
So what’s in store for the next 15 years? Will the industry undergo another fundamental shift in its landscape? Yes and no. The great leap forward to mobility — the uprooting of services and technology that once were confined to a specific place — already has become a given. The expectation is that every new application, every new service will now have — or will soon have — some kind of mobile component. Wireline voice connections are giving way to wireless, fewer and fewer computers are connected to networks via cords, and the Apple iPhone’s Safari browser already has begun to handle a noticeable percentage of the world’s Web browsing activity. Mobility’s already out of the bag.
Those trends will continue, of course. Networks will get faster, phones will get smarter and radio connections will reach a multitude of different devices. The next revolution won’t be about mobility per se, but it will use mobility to create unimaginable scale and depth in the networks of the future. The mobile operators of 2025 won’t be counting their subscribers in millions, but managing individual connections numbering in the billions. Any device or object with active electronics — and many without — will be capable of transmitting a wireless signal, communicating not just with the wide-area cellular network, but with multiple local and personal area networks, and often directly with one another. While device interfaces will evolve, a good deal of the traffic and transactions on the mobile networks will be between machines. Billions of sensors will flash data across networks, triggering countless numbers of interactions, most of which customers will have only a vague awareness of.
Devices will not only connect to the network, but interconnect directly with hundreds of different devices in the home, office, car and public space — some computing devices in their own right, others ordinary household objects. Wireless won’t just be a means of accessing the Internet. It will coax the Internet from the virtual world into the real one.
Russ McGuire, vice president of corporate strategy for Sprint, wrote a book called The Power of Mobility, the central premise of which is that the value of any object, application or idea increases relative to its mobility. The principle can be applied to almost any scenario: A famous work of art that is moved from one museum to another can be viewed by more people, thus increasing its aesthetic value to the whole world. A computer, a phone, even a business becomes more useful the less it is physically constrained.
McGuire, however, admits that mobility isn’t everything. Those devices have to work in unison, not as individual actors. Many of the technologies and applications that will inform the wireless world in 2025 are with us today, McGuire said. The key is connecting them in meaningful way.
“Let’s say I have an appointment in half an hour with Joe Smith of Galacticom Enterprises,” McGuire used as an example. “So I go and get in my car. My mobile device, of course, has my calendar, my address book and everything. As soon as I get in my car, my mobile device starts talking to other specialty devices in the car — maybe the navigation system, maybe the car radio — and they start sharing information. Immediately as I start the car, the car radio comes on, and it might ask me ‘What would you like to hear?’ Knowing I’m meeting with Joe Smith, it might ask me ‘Would you like to hear Joe’s 10 most recent status updates from Facebook, or would you like to hear the latest quarterly earnings results from Galacticom?’ It would tap into information available in the cloud based on information it just received from my mobile device on where I’m going. At the same time the navigation system pops up and says, ‘Do you want me to route you to Galacticom Enterprises?’
“As I drive, these pieces are working together to inform me and prepare me for the meeting I’m about to have. Perhaps when I’m halfway there the navigation system comes on and says, ‘Based on traffic, you’re probably going to be about 15 minutes late for your meeting. Do you want to call Joe and let him know?’ It interacts with the mobile device to start that call, probably working through a Bluetooth connection to the audio system. These three pieces plus the cloud all work in a way that’s synchronized, integrated and, I’d say, pretty different from how our world works today. It’s an experience that’s seamless, transparent and natural, yet significantly enhancing how we interact with the world.”
Just because two devices are linked together doesn’t mean they know what information to share. In McGuire’s example, it isn’t enough for the car’s onboard computer to have access to the phone’s calendar and address book. It must intuitively know which appointment is pertinent. It has to distinguish the Joe Smith, CEO of Galacticom, from the thousands of other Joe Smiths in any given American city. It has to understand that the meeting with Joe Smith is a face-to-face meeting, not a conference call. It has to reach out to the Internet cloud and communicate with dozens of databases, and then it has to filter through reams of data for relevance. The navigation system in the car not only has to retrieve a destination from the phone and access real-time traffic data, it has to understand that McGuire is on his way to an important appointment, not running an errand.
All of that information would be easily accessible today with a Web browser and, say, 15 minutes of research time. That information in turn could be tediously entered into those devices, which then could be programmed with safeguards and alerts. But the technology that could intuitively link all of those disparate bits of data together still isn’t in place. What’s needed is a “social grass” that identifies the relationships between those now-stranded bits of information; has the intelligence to contextualize it and decipher what pieces are relevant at any particular time, place or situation; and once filtered, synchronize it with dozens of devices over multiple networks, McGuire said. Much of the innovative effort of the next decade will be spent growing that grass.
THE SEMANTIC WEB AND THE INVISIBLE INTERNET
Wireless and Internet are converging, but not just in the obvious sense of the former being used to access the latter. Today devices are means of accessing content and information from the Web, but as they become more sophisticated, they start to become active nodes on the Internet, collecting data about themselves and their owners and sharing it with other devices and the network.
The World Wide Web Consortium is actively developing what it calls the “Semantic Web” — the first step in creating the framework McGuire said can tie together the disparate bits of isolated information on the Web today. While Web 2.0 accomplishes some of these goals by viewing the Web as a platform that can pull together resources and applications from various places and link them together, the Semantic Web would go further, creating an Internet meta-language that automatically would create associations between information, rather than depend on developers to manually design the interfaces between apps.
The Semantic Web would allow computers to analyze the whole of the Internet, accessing the meaning and context of any particular piece of content, rather than just the raw content itself. In McGuire’s car example, a Semantic Web networking application could use the limited information it has on Joe Smith from McGuire’s address book and calendar to determine his relationship with McGuire and his identity on the Web. The app would then search for Web information on Smith, which — because of his role as CEO of a corporation — likely would generate thousands of references. But because of the semantic data now embedded within the Web, the app could differentiate between an earnings report and a press release, between a blog that mentions Smith and a blog entry Smith created, even between a picture Smith posted and a picture of Smith. The app could then compare that semantic data against the relationship it has determined exists between Smith and McGuire (business partners), the context of their interaction (upcoming meeting) and McGuire’s current context (driving to said meeting), thus determining what information to share.
In that example, wireless permeates. It’s used as a connection technology to link the phone to the elements of the car and as an access technology to the Internet cloud. The wireless devices aren’t just passive receivers of content, either. Each device is sharing data about its user and its appointments with the network, and the car is updating its location. But the potential of wireless in such a semantic network ultimately could be far greater. As more devices become connected to the Internet, the combined information they can relay to individuals exponentially grows — far more than the average individual can reasonably interact with in a day. Researchers are starting to postulate the creation of an “Invisible Internet,” where devices no longer merely receive the information we use to make decisions, but use it to make their own decisions.
“When the number of connections outstrips the number of people in the world, you have hyper-connectivity,” said Vish Nandlall, carrier group chief technology officer and distinguished member of the technical staff for Nortel Networks. “But the degree to which, I think, the number of connections will outstrip the number of people in the world is staggering. As more and more things become connected, the information they collectively generate becomes more and more valuable. As that information becomes more and more valuable, I need to be able to process that information so I can get more and more intelligence out of it. Once I can do that, I get the Invisible Web.”
Nandlall said the Invisible Internet combines the intelligence of the Semantic Web with the inherent mobility of wireless, which suddenly allows all of that useful information to be acted upon in real time. For example, Nandlall said, if someone drove into a dangerous neighborhood, his GPS-enabled device would take note, immediately triggering network-based protocols that could lock the car doors or, if the car stopped moving for an extended period of time, dial 911. In another scenario, a museum patron might admire a work of art for a lengthy period of time. Sensors in her device could correlate which particular work she was viewing based not only on her precise location, but what direction she was facing or whether she was having a conversation with someone who happened to be in front of the artwork. If the patron’s viewing time passed a certain threshold, the device could deduce that she admired the piece and then signal back to the network to gather information on the artist, which would be downloaded to her e-book reader, or collect digital photographs of the artwork to be uploaded into a picture frame at her home or office.
The Invisible Internet is associated closely with the concept of the “Internet of Things,” in which a multitude of everyday objects are connected wirelessly. In such a world, not every object will have the intelligence to make decisions for itself — your carton of milk doesn’t need an advanced processor, only the ability to communicate what it is and its expiration date — but collectively they’ll create a form of ambient intelligence, allowing them to self-organize as a group. If the Invisible Internet of Things does become a reality, the Web will cease to be merely a virtual space, where people interact with one another from behind a PC or phone’s screen, and become a real space — “meat space,” if you will — where thousands of objects, both personal and public, interact with one another.
The one element, besides a radio, all of those objects have in common is awareness. They have to be able to sense one another as well as their surroundings. Embedding devices and objects with that kind of sensitivity probably is the smallest challenge the Internet of Things faces right now, said Henry Tirri, head of the Nokia Research Center. The core sensors needed in the network of the future already are embedded in the average smartphone today: GPS and cellular triangulation sense location; accelerometers and digital compasses sense movement and direction; digital cameras can see for the devices. Some of those sensors need to be refined, but for the most part, devices already have access to enormous amounts of raw sensory data, Tirri said. The challenge for the industry is processing that data, interpreting it and combining it with data from other sensors to make it useful. Once the technology overcomes those problems, there’s no limit to what can be wirelessly enabled, he added.
“In today’s world of handsets, we talk in billions; in the future, we will talk about trillions of devices,” Tirri said. “Radios and sensors will be very small. They will be in everyday devices like coffeemakers and key chains, as well as all consumer devices, but also things you wouldn’t think you’d have wireless capabilities, like chairs, tables, even your bed.”
Applying radios to furniture may sound far-fetched, but a few companies already have latched onto the concept, connecting ordinary household objects to the network. Violet sells radio-frequency ID (RFID) stamps that can be affixed to any ordinary object and then programmed to trigger any number of applications when scanned by Violet’s home RFID reader. For instance, a keychain, when flashed over the reader, immediately could prompt your PC to pull up traffic and weather information while simultaneously posting a status update on a social-networking site saying, “I’m leaving the house.” Another company, Ambient Devices, has designed an umbrella, ordinary in all ways but one: A radio is embedded in the handle that causes it to change color when there’s rain in the forecast.
“Any object that exists out there can be connected,” Nandlall said. “When that happens, as that density occurs in the connected universe, I think you’ll find this notion of the intelligent or the Invisible Internet will become manifest.”
Billions of devices on the network — some sending raw sensor data, others accessing high-bandwidth applications, but all of them interconnected — will place enormous strains on the network. All of those connections must be authenticated, all of the individual subscriptions managed and billed. As devices become more aware, they’ll constantly be letting the network and their peers know of their status — a flood of presence information. The biggest challenge, however, likely will be one of capacity. New mobile WiMAX and long-term evolution (LTE) networks will deliver to a single cell upward of 50 Mb/s of capacity over a 5 MHz downlink. That’s nothing to scoff at, but if wireless blossoms the way many in the industry are hoping and predicting, it won’t be enough to support the sheer volume of broadband connections necessary — at least not in today’s macro-cellular configurations.
The wireless industry traditionally has grown wireless capacity in one of two ways: increasing the efficiency of the radio interface — packing more bits into a hertz of spectrum — and finding new methods of reusing the spectrum it has — shrinking and splitting cells. Many of the great leaps forward in cellular technology have been mirrored by equally impressive innovations in spectral efficiency. The evolution from 2G to 3G boosted the capacity of the network to a point where the mobile industry could embrace data communications. We’re starting to see the next step in that evolution, as the first 4G networks go live, allowing wireless to support true broadband speeds.
While further advances in spectral efficiency are possible, great leaps no longer are. Wireless access technology has begun to brush up against a fundamental barrier in theoretical physics, called Shannon’s Law. Named after legendary Bell Labs scientist Claude Shannon, the law establishes a fundamental limit to the amount of error-free data that can be sent over a radio channel before any gains in capacity succumb to noise. With the introduction of orthogonal frequency division multiplexing in the WiMAX and LTE standards, wireless is hitting that limit. To feed the growing hunger for mobile broadband, the wireless industry will have to look for its capacity elsewhere.
The second technique, spectrum reuse, will become the prime way of eking out higher capacities from the network, and that won’t just mean the further splitting of cells. The cellular network will begin to shrink as bandwidth demands grows, said Håkan Eriksson, chief technology officer for Ericsson. The notion of the hulking cell site covering miles of urban terrain will evolve to one of smaller cells with centers spaced as closely as dozens of meters apart. A large city would be blanketed with hundreds of thousands of tiny cells — as opposed to hundreds of cells in the typical cellular deployment — each transmitting with the full capacity of a macro-cellular base station.
“Think of the macro-diversity encountered when simply driving down the street,” Eriksson said. “You would be passing through a new cell at every lamppost. There has to be tremendous mobility in such a solution.” The typical mobility management scenario of a single user occupying a single cell goes out the window, Eriksson said. A device would be in simultaneous contact with multiple cells, and the network would have to anticipate handoffs four or five cells down the chain.
The shrinking of the network already has begun occur with the introduction of the femtocell, which provides the equivalent of a personal site for an individual, business or family. But femtocells today are being used to fill in dead spots or offload traffic from the macro-network in a particularly high-use area. They haven’t been used to fundamentally change the architecture of networks. According to Michael Oommen, vice president of device and technology development for Sprint, femtocells will play a big part in future networks, but it would be a mistake to think of it in terms of just big and small cells. The networks of the future will require radically different topologies, Oommen said, many of which wouldn’t necessarily separate the notion of device from the notion of infrastructure. The devices themselves would become nodes in larger network topologies, communicating with one another along with multiple networks.
“The network is able to align itself to the needs of a customer respective to the device the customer has,” Oommen said. “I would say instead of a single network, you will have a hierarchical system of mini-networks. I wouldn’t call it macro, micro or even femto. Your device itself wouldn’t just be peer-to-peer, but peer-to-many or many-to-many. Your device should be able to sense the other devices around it. One device may be using another device’s capability to communicate or to do certain processing. I wouldn’t call that a femto network or a macro-network. I’d say the network itself becomes virtual.”
Read more here.
April 30th, 2009