2-M2M From mobile to embedded internet

I NTRODUCTION

Machine-to-machine (M2M) communications in

the context of the mobile Internet has been a

subject of intense discussions over the past two

years. Some see it as the next technology revolu-

tion after the computer and Internet. Some con-

sider it simply hype. Others are cautious with a

wait-and-see attitude.

Part of the confusion has been that M2M is

not something completely new. For those famil-

iar with embedded control, M2M is a natural

extension of their existing business. They fail to

see the explosive growth that others are excited

about. Other people also remember the high-

tech bubbles in recent history, and question the

practical future of M2M.

Intel recently completed an extensive study

on the issues critical to the M2M industry. We

exchanged views with leading equipment manu-

facturers, software vendors, and service pro-

viders. In this article, we share some of our

learning.

We begin with our vision of the future embed-

ded mobile Internet. Then we look at several

M2M use cases that offer significant market

potential. We discuss the requirements and chal-

lenges associated with mass-scale M2M networks,

and describe potential system architectures and

deployment options that can enable the connec-

tivity of billions of low-cost devices. We describe

the salient features of M2M traffic that may not

be supported efficiently by current standards and

provides an overview of potential enhancements.

Finally, we summarize the progress of standards

development for M2M.

T HE F UTURE OF

E MBEDDED I NTERNET

Mobile Internet is at a turning point. In this sec-

tion, we discuss what motivates the evolution

and share our vision of M2M for the future

embedded Internet.

T HE T ECHNOLOGY AND

E CONOMIC M OTIVATIONS FOR M2M

The proliferation of mobile Internet provides

nationwide ubiquitous coverage and mobility

support. Today’s advanced wireless networks are

ready to deliver broadband data service at a sig-

nificantly lower cost than in the past, thanks to

extensive standardization [1]. These networks

offer many of the features necessary to enable

M2M services in the future embedded Internet.

Technology is one of the main drivers of

M2M. The semiconductor industry’s shrinking

lithography and improved yield continue to

reduce chipset cost and power consumption.

Carrier WiFi, small cells, relay, and peer-to-peer

communication further extend the coverage of

wireless networks while dramatically reducing

cost per bit transferred.

There are also profound economic motiva-

tions for the wireless industry to aggressively

pursue M2M. As voice revenue continues to

deteriorate, operators are under tremendous

pressure to introduce new services that will fill

their revenue gap. M2M, cloud computing, and

application stores top the list of potential rev-

enue-generating services.

T HE V ISION OF I NTERNET OF T HINGS

To the authors, M2M represents a future where

billions to trillions of everyday objects and the

surrounding environment are connected and

managed through a range of devices, communi-

cation networks, and cloud-based servers.

There are three essential components to this

“Internet of Things” vision.

A BSTRACT

Is M2M hype or the future of our informa-

tion society? What does it take to turn the M2M

vision into reality? In this article we discuss the

business motivations and technology challenges

for machine-to-machine communications. We

highlight key M2M application requirements and

major technology gaps. We analyze the future

directions of air interface technology improve-

ments and network architectures evolution to

enable the mass deployment of M2M services. In

particular, we consider the salient features of

M2M traffic that may not be supported efficient-

ly by present standards, and provide an overview

of potential enhancements. Finally, we discuss

standards development for M2M.

Geng Wu, Shilpa Talwar, Kerstin Johnsson, Nageen Himayat, and Kevin D. Johnson, Intel

M2M: From Mobile to

Embedded Internet

A continuum of devices from low-cost/low-power to compute-rich/high-performance:In the M2M market, large numbers of devices are expected to be embedded, requiring extremely low price points and low power consumption.However, higher-end devices such as gateways,machine control modules, intelligent vision sys-tems, and even consumer electronic products are also important growth segments.

Ultra scalable connectivity:This is arguably the most important component of the M2M vision. A device that is not connected cannot easily be managed or work in concert with other devices. Our most critical challenge, therefore, is to enable low-cost connectivity that addresses not only the massive network scale but also the vastly diverse requirements dictated by the device continuum.

Cloud-based mass device management and services:The vision of the future is no longer one device acting alone, but many devices acting together. Thus, although distributed processing is critical to address the complexity of M2M

applications, centralized decision making and management of billions of devices within the

cloud will become an essential value of the Inter-net of Things vision.

Although this vision is not new, it is only now gathering momentum as ubiquitous connectivity is finally becoming a reality, and Moore’s Law has driven device cost and size low enough to justify “smart devices everywhere.”

T HE E SSENTIAL E LEMENTS OF

M2M S OLUTIONS

Third-/fourth-generation (3G/4G) wireless tech-nologies will play a central role in the future of M2M. Its high data rate enables high value ser-vices. In markets where 2G is reaching the end of its life cycle, 3G/4G is the only option.

As the M2M market expands, operators will encounter significant technical challenges. Secu-rity will be of paramount concern. A major secu-rity breach in a network connecting billions of devices is unthinkable. It is expected that advanced solutions including “security-on-chip”will be developed.

As the M2M market expands, “zero-touch”manageability and information overload will pose significant challenges to the network. We have observed in large-scale surveillance networks that the number of video feeds overwhelms human operators. Thus, there will be an urgent need for video analytics at the installed end devices.

Similarly, as M2M solutions evolve, optimum distribution of device and cloud intelligence will become critical. With increased intelligence at the device, augmented sensing will be practical for innovative value-added services.

Scaling smart device installations and sup-porting future technology is a crucial smart sys-tem architecture concern. We envision standardized M2M plug-n-play capability to be essential to overall acceptance of M2M tech-nologies.

Finally, future M2M solutions will need to support a mix of legacy and new services and devices . We expect M2M gateway/aggregation

points to play a key role in bringing the installed short-range sensors online and providing inter-working with different wireless technologies.These gateway/aggregation points can also become a platform for value-added services to enable an explosive growth of short-range smart sensors, fully managing the scale

T OWARD THE F UTURE OF E MBEDDED I NTERNET

Future M2M ecosystems will be complex and span many industries, including telecom and electronics. Unlike current M2M markets, which are highly segmented and often rely on propri-etary solutions, future M2M markets will need to be based on industry standards to achieve explosive growth. This standards process will be much broader than writing a specification, as it involves not only interfaces, but also platforms and services. The M2M industry needs to lever-age existing vertical market solutions, design platforms that horizontalize the market, and avoid the narrow solutions that come from chas-ing “killer apps.” The industry also needs to ramp up efforts to develop critical technologies for an optimized air interface, device manage-ability, network architecture, and security in order to enable future mass deployment of embedded devices.

R EPRESENTATIVE M2M U SAGE M ODELS

One can envision creating an immensely rich set of applications when thousands of objects sur-rounding us are connected. Some examples are smart homes, where intelligent appliances autonomously minimize energy use and cost;“connected cars” that react in real time to pre-vent accidents; and body area networks that track vital signs and trigger emergency response when life is at risk. In this section, we study three M2M applications and provide a brief description of other applications to demonstrate the broad market potential of M2M.

U TILITIES (S MART G RID )

Smart grid integrates communication capabilities with utility generation (e.g., electric power, gas,water) and delivery infrastructure to automate monitoring and control. Sgnificant savings in resource consumption is also possible when utility supply is dynamically matched with demand. Key smart grid applications are smart metering, distri-bution network automation, demand response,equipment diagnostics, as well as wide area moni-toring and control. An example of smart grid net-work architecture is described in [2].

An M2M-enabled smart meter collects utility usage information from home appliances via short-range radio or a home area network and sends the information to the M2M server by com-municating directly through the 3G/4G network or via an M2M aggregation device and then to the 3G/4G network. The M2M aggregation device collects information from many smart meters in the area and sends the aggregated information to the M2M application server. The home area net-work interface to the smart-meter can be based on several short-range wireless technologies such

A CKNOWLEDGMENT

The authors would like to thank Geoff Weaver and Boyd Bangerter of Intel Labs for their tech-nology insights and valuable comments.

R EFERENCES

[1] Morgan Stanley, “Internet Trends,” March 9, 2010

[2] IEEE C80216-10_0002r7, “Machine to Machine (M2M)

Communication Study Report,” IEEE802.16 Contribu-tion, May, 2010.

[3] DRAFT-T31-127-R020-v01, “Recommendations and

Requirements for WiMAX Machine to Machine (M2M), WiMAX Forum doc., Aug. 2010.

[4] 3GPP TS 22.368 v10.1.0, “Service Requirements for

Machine-Type Communications (Stage 1),” Release 10, June 2010.

[5] Cisco Visual Networking Index, Cisco VNI, Oct. 2009.

[6] IEEE C80216-10_0016r1, “Future 802.16 Network: Chal-

lenges and Possibilities,” Mar. 2010.

[7] S. Yeh, S. Talwar, S. Lee and H. Kim, “WiMAX Femto-

cells: A Perspective on Network Architecture, Capacity and Coverage,” IEEE Commun. Mag., Oct. 2008.

[8] Johansson et al.

, “A Methodology for Estimating Cost

and Performance of Heterogeneous Wireless Access

Networks,” PIMRC ’07.

B IOGRAPHIES

G ENG W U(geng.wu@https://www.wendangku.net/doc/7715954319.html,) is chief architect and direc-

tor of Wireless Standards in the Wireless Technology Divi-

sion at Intel Corporation. He has 20 years of experience in

the wireless industry. Prior to Intel, he was director of

Wireless Architecture and Standards at Nortel Networks,

with extensive experience in 3G/4G technology develop-

ments. He obtained his B.Sc. in electrical engineering from

Tianjin University, China, and his Ph.D. in telecommunica-

tions from Université Laval, Canada.

S HILPA T ALWAR(shilpa.talwar@https://www.wendangku.net/doc/7715954319.html,) is a principal

engineer in the Wireless Communications Laboratory at

Intel, where she is conducting research on mobile broad-

band technologies. She has over 15 years of experience

in wireless. Prior to Intel, she held several senior techni-

cal positions in the wireless industry. She graduated

from Stanford University in 1996 with a Ph.D. in applied

mathematics and an M.S. in electrical engineering. She is

the author of numerous technical publications and

patents.

K ERSTIN J OHNSSON(kerstin.johnsson@https://www.wendangku.net/doc/7715954319.html,) is a senior

research scientist in the Wireless Communications Labora-

tory at Intel, where she conducts research on network,

MAC, and PHY optimizations that improve wireless net-

work cost, coverage, and capacity. She graduated from

Stanford with a Ph.D. in electrical engineering and has

almost 10 years’ experience in the wireless industry. She is

the author of numerous publications and patents in the

field of wireless communications.

N AGEEN H IMAYAT(nageen.himayat@https://www.wendangku.net/doc/7715954319.html,) is a senior

research scientist with Intel Labs, where she performs

research on broadband wireless systems, including hetero-

geneous networks, cross-layer radio resource management,

MIMO-OFDM techniques, and optimizations for M2M com-

munications. She has over 15 years of research and devel-

opment experience in the telecom industry. She obtained

her B.S.E.E from Rice University and her Ph.D. in electrical

engineering from the University of Pennsylvania in 1989

and 1994, respectively.

K EVIN D. J OHNSON(kevin.d.johnson@https://www.wendangku.net/doc/7715954319.html,) is director

of Embedded Connected Devices at Intel Corporation. He

directs the team that has responsibility for accelerating

embedded products and technologies in the M2M indus-

try. He has over 20 years of product, technology, and

business experience in the computing and embedded mar-

kets. He holds a B.S. in engineering from Oregon State

University and a Master’s in business from the University

of Portland.