AT&T on their LTE Backhaul Architecture

Backhaul is a topic that may be giving some operators nightmare. Picked up this slightly old article from Light reading via WirelessMoves.

AT&T network architect Yiannis Argyropoulos addressed the Backhaul Strategies and Core Convergence for Mobile Operators event in New York City and had the following to say:

The lines between wireless and wireline networks are blurring, as are the boundaries between access and core networks, driven by the need to carry the flood of wireless data traffic more efficiently. AT&T is aggressively deploying fiber to its mobile cell sites and migrating from Sonet to Ethernet, but more changes will be needed. AT&T started its fiber push in 2008, and it will take at least seven years to complete, said Argyropoulos.

For the short term, today's metro Ethernet architecture will support LTE, but longer term, the network architecture needs to have less operational complexity, noted the AT&T man. The carrier is in the process of testing new approaches, based in part on work being done by 3rd Generation Partnership Project (3GPP) and the Broadband Forum .

AT&T also is looking for coordination of policy control between its wireline and wireless networks, so that the core network services are the same for end-users, regardless of how they connect to the network. It is no longer adequate for quality-of-service to be delivered piecemeal, within different segments of the network, Argyropoulos stated: "There is a lot of work going on right now to harmonize these."

The early 3GPP scheme for QoS on 3G UMTS networks was too complicated to be implemented, but newer LTE QoS plans from the 3GPP, with nine QoS classes and a smaller number of individual class attributes, look more practical.




The growing volume of data traffic is having an impact on other areas of the carrier's operations, too. The widespread use of bandwidth-hungry smartphone devices is creating new traffic patterns that, among other things, eliminate traditional maintenance windows traditionally scheduled in the early hours of weekend mornings, Argyropoulos pointed out.

"Data traffic peaks at the same time as voice, but it has multiple peaks, and it doesn't ever really subside," he said. That, in turn, is putting pressure on wireless network operators and their vendors to do hitless network upgrades and to build more resiliency into their networks.

AT&T is looking to other means of offloading traffic, including routing optimization that will use gateways strategically placed in the network to direct traffic onto the Internet, and not carry it through the metro and core networks first.

"Most of the mobile data traffic is coming from the Internet and going to the Internet."

It will also be important to offload subscriber traffic generated in the home onto a domestic Internet connection, he added.

To get an Idea of the mobile backhaul load, see my earlier post here.

Along with Fiber, Microwave is also an option and you can read more about it in Daily Wireless blog.

Also came across this blod dedicated to mobile backhaul, that is available here.

Ericsson Presentation: Cognitive radio in Europe

Cognitive radio in Europe [John Holland]

View more presentations from Keith.

'Femtocells' or 'Small cells' ?

Recently, while browsing, I ended up on Wilson Street. I have been noticing it since earlier this year that Alcatel-Lucent have rebranded their Femtocells as Small-cells. I have blogged earlier about Femtocell variations but the term 'small cell' could be used to cover different sizes and capacity of cells.

Here are some interesting things i found from a recent ABI research blog:

• Indoor residential grade Femtocells have an output power of 10mW-100mW.
• Enterprise grade or Metro femtocells have an output power of 200-300mW.
• Rural femtocells (a.k.a. Super Femtocells, Greater Femtocells) 200mW-1W. Some people refer to them as picocells as well.
• Compact base stations use femtocell silicon efficiencies and multi-core chipset platforms to build a base station on a SoC - but are meant to be higher output power base stations (1W and higher).
• Compact base stations are scalable platforms, which can fit into picocell, microcell or even macrocell form factors. The emergence of compact base station can be traced to the need for multifrequency, multimode, low power consumption, low-cost, pizza-box type base station platforms that can de deployed within different site classifications especially in metro metrozone overlays.
• The capacity crunch in networks is likely to drive operators to deploy compact base stations as in-fills initially with compact base stations being a part of future network blueprints. Current microcell or macrocell platforms are too bulky or costly to deploy in clusters and in large numbers. Compact base stations are also meant to take advantage of backhaul relay techniques making it easier to deploy in small clusters.
• Small cells on the other hand could be the umbrella under which compact base stations (portion of), picocells, microcells, residential, enterprise, rural/metro femtocells exist.
• We are already seeing vendors like Alcatel Lucent change their marketing message from femtocells to ‘small cells’ covering a wider range of products and deployment types. They have also included features like SON and value-added applications into the small cell base category.

To avoid confusing the end users who are just interested in better coverage and data rates, it would make sense to brand the Femtocell as something approprite (like Vodafone has done for Sure Signal). Small cells do sound good too.

The Indian entrepreneur spirit

In India you can fix and recycle anything and everything. The last time I visited India, I took an old Nokia phone whose screen and buttons were not working. The charger socket was damaged. There was no battery. The cover was damaged and there was nothing right in that. It did have sentimental value for someone who wanted it fixed if possible.

I gave it to someone to get it fixed and then when I saw it again after a week it was all nice and new. I has to pay 1400 Rupees (approx. £20/$30) but it was invaluable for the person who wanted it fixed.

I read a very interesting Blog article by Shekhar Kapur yesterday and I decided to share it with you. Shekhar Kapur is well known actor/director of movies and his well known international movies as director include Elizabeth and Bandit Queen.

You can read the complete blog article here.

Back in UK, I do nowadays see mobile repair shops springing up in different places but I still think they are far behind these Indian repair shops which dont have much equipment and components but can still fix your item miraculously.

I also found another Interesting blog that has some interesting articles on the theme of 'Indian Entrepreneur Spirit'. See here.

Coordinated Multi-Point (CoMP): Unresolved problems

I have blogged about CoMP in quite some detail in the past. Someone recently pointed out an interesting video from Fraunhofer Heinrich Hertz Institut which is embedded below:

CoMP may be not as practical as we may think. One of the things pointed out by Dr. Ariela Zeira, InterDigital's Vice-President of Advanced Air Interfaces in the LTE World Summit was that there exists a gap between the theoretical and the practical gains of CoMP.

She went on to suggest the following as way forward for the Coordinated Multipoint acceptance in future:

• Address root causes of gaps between academia and current feedback schemes
• Need for improved Channel State Information (CSI) feedback resolution
• Need for improved frequency domain precoding granularity
• Apply CoMP where most needed and/or theoretical gains can be approached
• Heterogeneous networks
• Interference problem is more severe than in macro-only deployment
• Especially for Femto Closed Subscriber Group and Pico Cells employing large cell extension
• Lower delay spread and low mobility can be expected in Femto and Pico cells and reduce performance loss from feedback impairments
• Relay Backhaul Channel (RBC)
• More accurate CSI feedback from stationary relay station is possible enabling advanced non-linear precoding schemes.
• High rank MIMO transmission will not be effective due to higher probability of Line of Sight (LOS) channel from Macro to Relay

CoMP is still probably the most promising spectral efficiency solution but need to focus on closing the gap between gains predicted by theory and those achievable with current LTE Release 8 Feedback Schemes

Challenges in Mobile phone 'Ad-Hoc' Networks

Last month I blogged about a Mobile phone based Ad-Hoc network. The following slide, again from Dr. Ariela Zeira, InterDigital's Vice-President of Advanced Air Interfaces shows the possible problems in having an LTE based approach for a Mobile phone 'Ad-Hoc' network.

This kind of technology is probably quite a few years away.

Double whammy for GSM Security

Via PC World:

A researcher at the Def Con security conference in Las Vegas demonstrated that he could impersonate a GSM cell tower and intercept mobile phone calls using only $1500 worth of equipment. The cost-effective solution brings mobile phone snooping to the masses, and raises some concerns for mobile phone security.

How does the GSM snooping work?

Chris Paget was able to patch together an IMSI (International Mobile Identity Subscriber) catcher device for about $1500. The IMSI catcher can be configured to impersonate a tower from a specific carrier. To GSM-based cell phones in the immediate area--the spoofed cell tower appears to be the strongest signal, so the devices connect to it, enabling the fake tower to intercept outbound calls from the cell phone.

What happens to the calls?

Calls are intercepted, but can be routed to the intended recipient so the attacker can listen in on, and/or record the conversation. To the real carrier, the cell phone appears to no longer be connected to the network, so inbound calls go directly to voicemail. Paget did clarify, though, that it's possible for an attacker to impersonate the intercepted device to the wireless network, enabling inbound calls to be intercepted as well.

But, aren't my calls encrypted?

Generally speaking, yes. However, the hacked IMSI catcher can simply turn the encryption off. According to Paget, the GSM standard specifies that users should be warned when encryption is disabled, but that is not the case for most cell phones. Paget explained "Even though the GSM spec requires it, this is a deliberate choice on the cell phone makers."

What wireless provider networks are affected?

Good news for Sprint and Verizon customers--those networks use CDMA technology rather than GSM, so cell phones on the Sprint or Verizon networks would not connect to a spoofed GSM tower. However, AT&T and T-Mobile--as well as most major carriers outside of the United States--rely on GSM.

Does 3G protect me from this hack?

This IMSI catcher hack will not work on 3G, but Paget explained that the 3G network could be knocked offline with a noise generator and an amplifier--equipment that Paget acquired for less than $1000. With the 3G network out of the way, most cell phones will revert to 2G to find a viable signal to connect to.

Another one from CNET:

A researcher released software at the Black Hat conference on Thursday designed to let people test whether their calls on mobile phones can be eavesdropped on.

The public availability of the software - dubbed Airprobe -- means that anyone with the right hardware can snoop on other peoples' calls unless the target telecom provider has deployed a patch that was standardized about two years ago by the GSMA, the trade association representing GSM (Global System for Mobile Communications) providers, including AT&T and T-Mobile in the U.S.

Most telecom providers have not patched their systems, said cryptography expert Karsten Nohl.

"This talk will be a reminder to this industry to please implement these security measures because now customers can test whether they've patched the system or not," he told CNET in an interview shortly before his presentation. "Now you can listen in on a strangers' phone calls with very little effort."

An earlier incarnation of Airprobe was incomplete so Nohl and others worked to make it usable, he said.

Airprobe offers the ability to record and decode GSM calls. When combined with a set of cryptographic tools called Kraken, which were released last week, "even encrypted calls and text messages can be decoded," he said.

To test phones for interception capability you need: the Airprobe software and a computer; a programmable radio for the computer, which costs about $1,000; access to cryptographic rainbow tables that provide the codes for cracking GSM crypto (another Nohl project); and the Kraken tool for cracking the A5/1 crypto used in GSM, Nohl said.

More information about the tool and the privacy issues is on the Security Research Labs Web site.

Interdigital's 'Fuzzy Cells' technology for cell edge performance improvement

Back in LTE World Summit 2010, I heard from Dr. Ariela Zeira, InterDigital's Vice-President of Advanced Air Interfaces about various things Interdigital have been working on.

One of the technologies that caught my attention was Fuzzy Cells technology to increase the cell edge rates. The following is from their press release for Mobile World Congress:

BARCELONA, Spain, Feb 15, 2010 (BUSINESS WIRE) -- InterDigital, Inc. today announced the demonstration of its "Fuzzy Cell" technology that improves cell-edge performance at the 2010 Mobile World Congress. The Fuzzy Cell technology is part of the company's comprehensive suite of "Next Generation Cellular" (NGC) innovations that combine advanced network topologies and spectrally-efficient air interface solutions for LTE-advanced and beyond.

"Many wireless operators and customers are experiencing a substantial degradation of service quality caused by the ever-growing demand for mobile data," said James J. Nolan, Executive Vice President, Research and Development, InterDigital. "We are at the forefront of developing solutions for more efficient wireless networks, a richer multimedia experience, and new mobile broadband capabilities that support operators to capture revenues from the boom in smartphones. The Fuzzy Cell fits nicely within our much broader efforts on spectrum optimization, cross-network connectivity and mobility, and intelligent data delivery techniques."

While cellular networks have become virtually ubiquitous, users continue to experience inconsistent and unpredictable performance when moving around. While this degradation is often the result of network congestion or an obstructed path of the radio waves, it is also inherent to traditional cellular deployments, whereby signals degrade towards the fringe of any given cell due to interference from neighboring cells. It is estimated that typical users experience this situation, known as being in the cell-edge, more than 50% of the time. Advancements in HSPA and LTE primarily increase peak data rates and only offer modest improvements in average performance throughout a cell.

Fuzzy Cells is a novel approach for leveraging existing resources to improve spectral efficiency and cell-edge performance. In a traditional deployment a device connects to one site at a time (even if multiple sub-bands are used at each site) and all sites use the same power levels and sector orientations for all sub-bands. In a Fuzzy Cell deployment, a device may connect to multiple sites at a time through the different sub-bands and continue to realize full system bandwidth. The power levels and sector orientations of the different sub-bands are optimized for best performance. In simpler terms, the device exploits the best combination of base station support regardless of its position, removing traditional limitations of cell or sector boundaries. Importantly, Fuzzy Cell technology can also allow gains indoors as it allows connection to more than one cell/sector at a time as available. The Fuzzy Cell technology provides additional improvement over Fractional Frequency Reuse (FFR) methods that are supported by current specifications.

The following shows the demonstration of Fuzzy cells:

I haven't heard any news recently on this technology but its an interesting concept, not sure if it would be adopted in the near term in the standards.

The Tester's Prayer

The Tester's Prayer

Oh Lord, give us our daily bugs;

and the wisdom to find the solution or pass the blame.

Oh Lord, help us find the problems before the developers;

for if they fix the problems there may be none for us left to find.

Oh Lord, make sure that developers dont fix their problems properly;

for it gives us time to sit and relax.

Oh Lord, grant us the patience when same problem comes again and again;

for some things are better left unfixed for us to find.

Oh Lord, grant us serenity when all things work fine;

and the belief that things will soon fall in line.

Thanks to everyone who contributed and helped refine this over the years :)

NTT DoCoMo announces 'Crossy'

NTT DoCoMo unveiled the brand name and logo of its forthcoming LTE next-generation mobile service for the Japanese market, which is scheduled to launch in December. The brand name is written Xi™ and read “crossy.” The logo is shown below.

The “X” denotes both “connection” and “infinite possibility,” and the “i” both “individual user” and “innovation.” The logo, which resembles the infinity symbol, aligns the letters in a single stream to embody the bonds that organically link people, goods and information, and lead to new innovation.

DOCOMO’s Xi LTE service will offer downlinks of up to 75 Mbps, approximately 10 times faster than the company’s current FOMA™ 3G service. Initially, Xi will be available in the Tokyo, Nagoya and Osaka areas, but coverage eventually will be expanded to other major cities and then additional areas of the nation. Xi users will be seamlessly handed over to the FOMA network whenever they leave a Xi service area.

Xi handsets, billing plans and other details will be announced later.

Meanwhile the world got its fourth commercial LTE network this week, and its first outside the Nordic region - in the unlikely telecoms hotspot of Uzbekistan.

The Uzbek launch came from Russian giant MTS, and shows how some emerging global telecoms players plan to leapfrog rivals by deploying LTE or WiMAX in markets where 3G is under-developed, and then harness that experience for their mainstream territories. MTS follows TeliaSonera's three LTE commercial switch-ons (in Sweden and Norway plus a soft launch in Finland).

MTS calls its new network "the first fully operational 4G network in the CIS and Central Asia" and is it initially available in the central zone of capital Tashkent, boasting theoretical peak speeds of up to 100Mbps using dongles. The network runs in the 2.5-2.7GHz band and the equipment comes from Huawei.