Evolution of Pervasive Computing

1INTRODUCTION

A web of computing devices and sensors embedded in everyday objects ranging from cars to house appliances. The last few years have seen a dramatic change in the role of technology in our lives. We now use the tools of information technology in multiple ways without even thinking about it. This phenomenal growth of the influence of technology can be expected to continue at this same mindboggling rate, and evolve into a form in which it is seamlessly integrated into our everyday lives. The advancement of computing and communication technology on these lines can be expected to have a profound effect similar to that which the discovery of electricity had in the development of our civilization. Computing is no longer a discrete activity bound to a desktop; network computing and mobile computing are fast becoming a part of everyday life and so is the Internet. Rather than being an infrastructure for computers and their users alone, it is now an infrastructure for everyone. The essence of Pervasive Computingis to create saturated environments with computing and wireless communications capability, yet gracefully integrated with human users. Many key building blocks needed for this vision are now viable commercial technologies: wearable and handheld computers, high bandwidth wireless communication, location sensing mechanisms, and so on .We expect devices like PDAs (Personal DigitalAssistants), mobile phones, office PCs and even home An entertainment system to access information and work together in one integrated system and the challenge is to combine these technologies into a Seamless whole and on the Internet. The aim of Pervasive Computing is for making computing available wherever it's needed. It spreads intelligence and connectivity to more or less everything. So conceptually, ships, aircrafts, cars, bridges, tunnels, machines, refrigerators, door handles, lighting fixtures, shoes, hats,packaging clothing, tools, appliances, homes and even things like our coffee mugs and even the human body will be embedded with chips to connect to an infinite network of other devices and to create an environment where the connectivity of devices is embedded in such a way that it is unobtrusive and always available. Pervasive Computing, therefore, refers to the emerging trend toward numerous, easily accessible computing devices connected to an increasingly ubiquitous network infrastructure. Pervasive Computing aims to make our lives simpler through the use of tools that allow us to manage information easily. These "tools" are a new class of intelligent, portable devices that allow the user to plug into Powerful networks and gain direct, simple, and secure access to both relevant information and services. Pervasive Computing devices are not personal computers as we tend to think of them, but very tiny - even invisible - devices, either mobile or embedded in almost any type of object imaginable; all communicating through increasingly interconnected networks. Information instantly accessible anywhere and anytime is what Pervasive Computing is all about!

1.1Comparison with traditional networking:

Communication between various devices isa central component of pervasivetechnology. However, in this it is differentfrom traditional networking as there are noindividuals directly involved in the processof communication. Rather, the devicesintelligently figure out the informationrequired and make the appropriateconnections by themselves, without userintervention. Thus these appliancescommunicate over networks such thatpeople do not directly monitor thecommunication between machines andprograms. The majority of thesecommunications will occur in an end to- endstructure that does not include a human atany point. This automatic operationillustrates the features of “smart” devices ofthe future. The number of machinesconnected to the Internet has been increasingat an exponential rate and will continue togrow at this rate as the existing networks ofembedded computers, including those thatalready exist within our automobiles, areconnected to the larger, global network, andas new networks of embedded devices areconstructed in our homes and offices. Thekinds of devices that will be used to accessthe Internet are no longer confined todesktops and servers, but include smalldevices with limited user interface facilities(such as cell phones and PDAs); wirelessdevices with limited bandwidth, computingpower, and electrical power; and embeddedprocessors with severelimitations on theamount of memory and computing poweravailable to them. Many of these devices aremobile, changing not only geographicposition, but also their place in the topologyof the network. Unlike traditional DesktopComputers and existing networks, the newdevices will have the followingcharacteristics:

• As described earlier, they will connect toother computing elements without the directintervention of users.
• The connections can be expected to bewireless in most cases.
• Many will have small, inexpensiveprocessors with limited memory and little orno persistent storage.
• The devices themselves will evolve veryrapidly. Old technologies will fail and newerones willappear much more frequently than atpresent.
• They will act as terminals for receivingand sending information.

1.2Advantages of pervasive computing:

The advantages of Pervasive Computingcanbe summarized as dramatically increasedautomation without much obtrusive orvisible use of computer infrastructure. Weincreasingly rely on the electronic creation,storage, and transmittal of personal,financial, and other confidential information,and demand the highest security for all thesetransactions and require complete access totime-sensitive data, regardless of physicallocation. We expect devices – personaldigital assistants, mobile phones, officePC and home entertainment systems --to access that information and work togetherin one seamless, integrated system.

Pervasive Computinggives us the tools tomanage information quickly, efficiently, andeffortlessly. It will help people makeeffective use of computing tools withoutrequiring the knowledge of complex devices, by means of a new class ofintelligent and portable appliances or "smartdevices" embedded with microprocessorsthat allow users to plug into intelligentnetworks and gain direct, simple, and secureaccess to both relevant information andservices. It gives people convenient accessto relevant information stored on powerfulnetworks, allowing them to easily takeaction anywhere, anytime. Pervasive Computingsimplifies life by combiningopen standards-based applications witheveryday activities. It removes thecomplexity of new technologies, enables usto be more efficient in our work and leavesus more leisure time and thus Pervasive Computingis fast becoming a part ofeveryday life.

1.3Evolution of Pervasive Computing

Pervasive Computingdefines a major evolutionary step in work that began in the mid 1970s, when the PC first brought computers closer to people. In Weiser’s vision, however, the idea of making a computer personal is technologically misplaced. In fact,it keeps computing separate from our daily life. Although the PC has not delivered the full potential of information technology to users, it certainly took a first step toward making computers (if not computing) popular (if not pervasive). It was also an instrumental factor in the phenomenal growth of hardware components and the development of graphical user interfaces.

Figure 1.1 System view of pervasive computing. The mobile computing goal of “anytime anywhere” connectivity is extended to “all the time everywhere” by integrating pervasiveness support technologies such as interoperability, scalability, smartness, and invisibility.

1.3.1Distributed computing

With the advent of networking, personal computing evolved to distributed computing. As computers became connected, they began to share capabilities over the network. Distributed computing marked the next step toward Pervasive Computingby introducing seamless access to remote information resources and communication with fault tolerance, high availability, and security. Although the World Wide Web was not designed to be a distributed computing infrastructure, its networking ubiquity has made it an attractive choice for experimenting with distributed computing concepts. It has also created a culture that is substantially more amenable to the deployment of Pervasive Computing environments than the culture that existed when Weiser first articulated his vision. The ad hoc nature of the Web’s growth has proved that we can distribute computing capabilities in a big way without losing scalability. The simple mechanisms for linking resources have provided a means for integrating distributed information bases into a single structure.

Most importantly, the Web has pioneered the creation of a nearly ubiquitous information and communications infrastructure. Many users now routinely refer to their point of presence within the digital world—typically, their homepages, portals, or e-mail addresses. The computer they use to access these “places” has become largely irrelevant. Although the Web does not pervade the real world of physical entities, it is nevertheless a potentialstarting point for Pervasive Computing.

1.3.2Mobile computing

Mobile computing emerged from the integrationof cellular technology with the Web.4 Both the sizeand price of mobile devices are falling everyday andcould eventually support Weiser’s vision of pervasiveinch-scale computing devices readily availableto users in any human environment. Cellular phonesystems that separate the handset from the subscriberidentity module (SIM) card approximatethis model of operation. Subscribers can insert theirSIM card and automatically use any handset, placingand receiving calls as if it were their own phone.Users can already access the same point in theWeb from several different devices—office or homePC, cell phone, personal digital assistant, and soforth. In this sense, for most users, what matters isthe view a particular machine provides of the digitalworld. SIM cards also demonstrate that the endsystem is becoming less important than the accessto the digital world. In this sense, we are well on

the way to computers “disappearing,” freeing usersto focus beyond them.The “anytime anywhere” goal of mobile computingis essentially a reactive approach to informationaccess, but it prepares the way for Pervasive Computingproactive “all the time everywhere”goal. As Figure 1 shows, pervasive computing is asuperset of mobile computing. In addition to mobility,pervasive systems require support for interoperability,

scalability, smartness, and invisibility toensure that users have seamless access to computingwhenever they need it.

2PERVASIVE COMPUTING MODEL

2.1Necessary technological advances

The technological advances necessary to build aPervasive Computingenvironment fall into fourbroad areas: devices, networking, middleware, andapplications. Figure 2 illustrates their relationships.

Figure 2.1 Pervasive computing framework. Middleware mediates interactions with the networking kernel on the user’s behalf and keeps users immersed in the pervasive computing space.

2.1.1Devices

An intelligent environment is likely to containmany different device types:

1. Traditional input devices, such as mice or keyboards,and output devices, such as speakersor light-emitting diodes;
2. Wireless mobile devices, such as pagers, per-sonal digital assistants, cell phones, palmtops,and so on; and
3. Smart devices, such as intelligent appliances,floor tiles with embedded sensors, and biosensors.

Ideally, Pervasive Computingshould encompassevery device worldwide that has built-in active andpassive intelligence. The University of Karlsruhe’sMediaCup project is an experimental deploymentof everyday objects activated in this sense. The project’s

guiding principle is to augment objects witha digital presence while preserving their originalappearance, purpose, and use.

Sensors that automatically gather information,transfer it, and take actions based on it represent animportant subset of pervasive devices. For example,sensors based on the Global Positioning System providelocation data that a device can translate into

an internal representation of latitude, longitude, andelevation. Stereo camera vision is another effectivesensor for tracking location and identity in a pervasiveenvironment. These fast-processing, two-lensdigital cameras can record both background imagesand background shapes. The results are much morerobust for tracking motion such as gestures.

2.1.2Pervasive networking

The number of pervasive devices is expected tomultiply rapidly over the next few years. IDC, amarket analysis firm, has predicted that, by the endof 2003, the number of pervasive devices willexceed the estimated worldwide population of 6billion people. Specifically, there will be more than300 million PDAs; two billion consumer electronic

devices, such as wireless phones, pagers, and set topboxes; and five billion additional everydaydevices, such as vending machines, refrigerators,and washing machines embedded with chips andconnected to a pervasive network.

As a consequence of this proliferation, many currenttechnologies must be revamped. In additionto extending the backbone infrastructure to meetthe anticipated demand, global networks like theInternet also must modify existing applications tocompletely integrate these Pervasive Computingdevices into existing social systems.

2.1.3Pervasive middleware

Like distributed computing and mobile computing,Pervasive Computingrequires a middleware“shell” to interface between the networking kerneland the end-user applications running on pervasivedevices. As Figure 2 shows, this pervasive middleware

will mediate interactions with the networkingkernel on the user’s behalf and will keep usersimmersed in the Pervasive Computingspace. Themiddleware will consist mostly of firmware andsoftware bundles executing in either client-serveror peer-to-peer mode.

User interfaces are another aspect of middleware.Standard Web browsers represent the high end ofinterface sophistication. They use more color,graphics, and controls than users typically expecton pervasive devices. Mobile computing hasalready introduced microbrowsers. For example,phone.com’s UP.Browser is implemented on severalcommercially available digital phones. “Middleware must mask heterogeneity to make pervasive computing invisible to users.”

2.1.4Pervasive applications

Pervasive Computingis more environment-centricthan either Web-based or mobile computing. Thismeans that applications will guide the middlewareand networking issues to a large extent.Consider a heart patient wearing an implantedmonitor that communicates wirelessly withcomputers trained to detect and report abnormalities.The monitor should know when toraise the alarm, based on its knowledge aboutthe environment. So this is much more thansimple wireless communication.

EliteCare ( is an assistedliving complex that applies similar technologies toimprove the quality of life for elderly residents.

3ISSUES AND CHALLENGES

As a superset of mobile computing, Pervasive Computingsubsumes mobile computing’s researchissues while opening up new ones unique to itself.In all cases, pervasive applications should disappearinto the environment.

3.1Issues and challenges in Pervasive Computing

Issues and challenges in pervasive are as described below:

3.1.1Scalability

Future Pervasive Computingenvironments willlikely face a proliferation of users, applications, networkeddevices, and their interactions on a scalenever experienced before. As environmental smartnessgrows, so will the number of devices connectedto the environment and the intensity of humanmachineinteractions.

Traditional development requires recreating theapplication for each new device. Even if an enterprisecould generate new applications as fast as it adds newdevices, writing application logic only once—independentof devices—would have tremendous value insolving the applications scalability problem.Furthermore, applications typically are distributedand installed separately for each device classand processor family. As the number of devicesgrows, explicitly distributing and installing applicationsfor each class and family will becomeunmanageable, especially across a wide geographicarea.

3.1.2Heterogeneity

Conversion from one domain to another is integralto computing and communication. Assumingthat uniform and compatible implementations ofsmart environments are not achievable, Pervasive Computingmust find ways to mask this heterogeneity—or uneven conditioning, as it has been called—from users. For instance, a sophisticated laboratoryand a department store may always differ in theirinfrastructural smartness. Pervasive Computingmustfill this gap at some level, say middleware, to smoothout “smartness jitter” in the user’s experience.

For networking, developers have faced protocolmismatch problems and learned how to tackle thelarge dynamic range of architectural incompatibilitiesto ensure transnetwork interoperability.Mobile computing has already achieved disconnectedoperation, thereby hiding the absence ofwireless coverage from the user. Middleware mayborrow similar concepts to dynamically compensatefor less smart or dumb environments so thatthe change is transparent to users.

But the real difficulty lies at the application front.Today, applications are typically developed for specificdevice classes or system platforms, leading toseparate versions of the same application for handhelds,desktops, and cluster-based servers. As heterogeneityincreases, developing applications thatrun across all platforms will become exceedinglydifficult.

3.1.3Integration

Though Pervasive Computingcomponents arealready deployed in many environments, integratingthem into a single platform is still a research problem.The problem is similar to what researchers indistributed computing face, but the scale is bigger.As the number of devices and applications increases,integration becomes more complex. For example,servers must handle thousands of concurrent clientconnections, and the influx of pervasive deviceswould quickly approach the host’s capacities. Weneed a confederation of autonomous servers cooperatingto provide user services.

Integrating Pervasive Computingcomponents hassevere reliability, quality of service, invisibility, andsecurity implications for pervasive networking. Theneed for useful coordination between confederationcomponents is obvious. This coordinationmight range from traditional areas such as messagerouting or arbitrating screen usage to new challengessuch as deciding which application can usea room’s light intensity to communicate with theuser.

3.1.4Invisibility

A system that requires minimal human interventionoffers a reasonable approximation of invisibility.Humans can intervene to tune smartenvironments when they fail to meet user expectationsautomatically. Such intervention might alsobe part of a continuous learning cycle for the environment. To meet user expectations continuously,however, the environment and the objects in it mustbe able to tune themselves without distracting usersat a conscious level.

A smart environment can implement tuning atdifferent system levels. For example, network-leveldevices will require autoconfiguration. Currentmanual techniques for configuring a device withaddresses, subnet masks, default gateways, and soon are too cumbersome and time-consuming forPervasive Computing.Automated techniques to dynamically reconfigurethe network when required are also crucial torealizing the Pervasive Computingvision.

3.1.5Perception: Context awareness

Most computing systems and devices today cannotsense their environments and therefore cannotmake timely, context-sensitive decisionsPervasive Computing, however, requires systems and devicesthat perceive context. Mobile computing addresseslocation- and mobility-management issues but in areactive context—responding to discrete events.