Tuesday, July 29, 2014

LTE network for US public safety taking it one step at a time

The organizers of the FirstNet LTE public safety network have the frequencies and standards they need to build the system, and they know where the money's coming from. They know how to get there from here, but it won't be a quick trip.

FirstNet will realize a vision that emerged in the wake of the 9/11 terror attacks, using technology that didn't exist until years later. It will be a single network linking all federal, state and local public-safety agencies in the U.S., based on the same radio spectrum and technology throughout. Though it won't replace every public-safety radio system in use today, FirstNet will help to eliminate the crazy quilt of incompatible radio systems and frequencies that makes it hard for different teams to coordinate their efforts.

That's no small matter when the news is bad enough to send first responders from multiple cities, counties or states converging on one area. For example, the many firefighting forces that battle summer blazes around the West often can't communicate directly with each other because they use different types of radios and different frequency bands, said TJ Kennedy, acting general manager of the First Responder Network Authority (FirstNet), which is in charge of making the network a reality.

The systems that first responders use now, including more than 10,000 separate LMRS (land mobile radio system) networks, also fall short of many users' needs. Some public-safety employees have to use their own smartphones in order to use apps, send photos and make calls in the field, according to Kennedy. Once FirstNet's built, all agencies will be able to sign up for the same national service, built on modern mobile broadband technology. It will span not just the 50 states but also U.S. territories, such as Puerto Rico, Guam and the Virgin Islands, and is intended to cover as much land as possible. In some cases that will probably require satellite, but most wireless will go over land-based LTE.

As with any effort to coordinate across 50 states and six territories, spanning about 60,000 public safety agencies, the network won't happen overnight. In fact, FirstNet isn't committing to any precise timeline or budget for getting it done. To give an idea how long the effort might take, there's a 46-step process that has to be carried out for each state and territory. The group is making progress: In many states, it's on step 7, Kennedy said.

That long process is designed to make sure the FirstNet system serves the needs of each state. FirstNet is meeting with local agencies and others involved with the issue, educating them about the technology and finding out what they want out of it.

"The geography and the needs of public safety in Maryland are probably very different from the needs in Alaska," Kennedy said.

Ultimately, each state and territory will choose whether to build the local wireless portion of the network themselves or have FirstNet do it. They can't opt out of the system altogether. Once the wireless infrastructure is in place, individual police departments, fire departments and other agencies will sign up and pay for service on FirstNet in much the same way they now buy service from a commercial mobile operator. FirstNet expects the service to be competitively priced, Kennedy said.

The network itself will be built and operated by carriers or other bidders that respond to FirstNet RFPs (requests for proposals), which will lay out the requirements for the system. Those criteria are still being set.

There's better news on the funding and technology for FirstNet.

Though not all the money is there yet, the funding sources for the system are secure, Kennedy said. The law that authorizes the network says the money to build it will come from three national auctions of wireless spectrum, which are forecast to bring in about US$7 billion. One of those, the so-called H Block auction, has already generated about $1.5 billion. Still to come are the sale of a band called AWS-3 to mobile operators, coming in November, and later the so-called incentive auctions to convert TV frequencies to mobile broadband.

FirstNet is also likely to be an easy fit with other networks and devices. It's designed to run entirely on IP (Internet Protocol), with a fast wired backbone in the core and LTE wireless networks at the edge. Because all the major commercial carriers in the U.S. use LTE, any gear that goes into the network or into first responders' hands can be based on the same mass-produced technologies, keeping costs down.

Unlike current public-safety systems, FirstNet will also have enough bandwidth to carry voice, video and data on mobile devices. The network has been assigned a 20MHz chunk of spectrum in the 700MHz band, comparable to what the major commercial carriers are using in that band. Carriers like 700MHz for its long-reaching signals and ability to penetrate walls.

Some devices on the market already are equipped to use FirstNet's band, and more will follow, Kennedy said. Some other countries have adopted the same band for public safety, most importantly Canada, which shares a continent-wide border with the U.S. This could allow for interoperability between U.S. and Canadian systems if needed, he said.

Stephen Lawson covers mobile, storage and networking technologies for The IDG News Service. Follow Stephen on Twitter at @sdlawsonmedia. Stephen's e-mail address is stephen_lawson@idg.com


View the original article here

Oracle's new in-memory database option could spark unanticipated costs, expert warns

Oracle database shops that have or are planning to download the latest version of 12c take warning: The vendor's newly launched, much-hyped in-memory processing database option is turned on by default, according to one expert.

The in-memory option costs US$23,000 per processor, according to an Oracle price list updated this week. Customers who don't realize the option has been switched on may find their next license audit "um, more entertaining," wrote Kevin Closson, a senior director in EMC's performance engineering group and a former Oracle architect who worked on its Exadata database machine, in a post on his personal blog this week.

"Please let me point out that I'm trying as hard as I can to not make a mountain out of a molehill," Closson wrote. The new version of database 12c containing the in-memory option is "hugely important," he added.

However, given the "crushing cost of this option/feature I expect that its use will be very selective," he wrote. "It's for this reason I wanted to draw to people's attention the fact that -- in my assessment -- this option/feature is very easy to use 'accidentally.' It really should have a default initialization setting that renders the option/feature nascent -- but the reality is quite the opposite."

An Oracle spokeswoman did not respond to a request for comment Friday on Closson's blog post, which was first highlighted by The Register.

Opinions on whether the option's cost is truly "crushing" may vary. Other Oracle database options, such as Real Application Clusters, are priced similarly. Oracle customers can typically negotiate significant discounts off list prices as well, although such a discussion may not be possible if any misuse of the in-memory option, accidental or not, is discovered during an audit.

Closson's blog post sparked a press release from Mark Flynn, CEO of the nonprofit organization Campaign for Clear Licensing. The group is lobbying software vendors in hopes of making licensing terms clearer.

"Oracle quite rightly deserves to make a lot of money from this innovation, but we fear that a large proportion of the additional income that it will generate (particularly in the short-term) will be through end-users being stung at their next audit because they were not aware of the change," Flynn said in the release.

Ultimately, database administrators are responsible for making sure their systems are license-compliant, Flynn noted. However, "we do not live in a perfect world," he said. "Admins have a million and one other priorities in their day -- keeping up-to-date with the latest licensing changes is rarely top of their list."

The onus is on vendors such as Oracle to better educate customers when changes like this are made, he added.

There's no question Oracle wants a large number of its database customers to use the in-memory option, which it's using as a hedge against customer defections to rival in-memory platforms from SAP, Microsoft and IBM.

Oracle's approach creates an in-memory column store, which dramatically speeds up analytic queries, while preserving the database's existing relational row store for OLTP (online-transaction-processing) workloads. The column store mitigates the overhead required to maintain row-based analytic indexes, improving OLTP performance.


View the original article here

EU, Google, Microsoft, Yahoo meet on 'right to be forgotten' but questions remain

European data protection authorities still have questions after meeting with Google, Microsoft and Yahoo about the implementation of a recent ruling that gave European citizens the right to be forgotten by search engines.

The search engine providers have until the end of the month to answer additional questions in writing.

Google, Microsoft and Yahoo met with EU data protection authorities who are members of the EU's Article 29 Working Party (A29WP) in Brussels on Thursday to discuss the May ruling by the Court of Justice of the European Union (CJEU). The ruling gave people the right to compel search engines to remove search results in Europe for queries that include a person's name, if the results shown are "inadequate, irrelevant or no longer relevant, or excessive."

However, the implementation of the ruling has turned out to be difficult to execute. Google already described the guidelines for removing query results "very vague and subjective."

The A29WP data protection authorities (DPAs) said in a news release that they had requested the meeting with the search engine officials in order to get input for future guidelines. The aim is to ensure a consistent implementation of the take-down ruling on the part of the search engine providers as well as consistent handling of complaints lodged with the authorities by people whose requests were denied, the DPAs said.

Confusion about the ruling could lead to a large number of complaints that the DPAs would then have to deal with -- a situation they apparently want to avoid.

Google said at the meeting that it has refused about 30 percent of requests, according to the statement from the DPAs. So far, the search engine has received 91,000 take-down requests concerning 328,000 links to Web addresses, a Google spokesman confirmed. About 15 percent of requests prompted Google to ask additional information. Over half of all requests have been granted.

Other than confirming take-down figures, Google declined to comment on the meeting, as did Microsoft. Yahoo did not immediately respond to a request for comment.

During the meeting the DPAs asked the search engines to explain their delisting process, according to the news release. The DPAs said they asked what criteria search engines use when balancing their economic interest and the interest of the general public in having access to information with the right of the person who wants the search results delisted.

Search engines were also asked if they notify website publishers when links are removed -- something that Google does -- and if so, what legal basis they have for sending out notifications.

The DPAs also wanted to know on what domains search results are delisted. Google for instance removes results from its European domains but not from its .com domain. It argues that the .com domain is not covered by the ruling because it is not aimed at Europeans in particular.

Time seemed to have been too short on Thursday to discuss all the DPAs' concerns. Google, Yahoo and Microsoft were asked to answer additional questions in writing by July 31. By then, they should provide details about the proof of identify or authentication they demand from people who file take-down requests. They also were asked to describe what safeguards are in place to protect any personal data processed during the handling of delisting requests.

In addition, the search engines were asked whether they post notifications on search results pages letting users know when some results have been removed due to EU law -- which is something that Google does -- and asked what the legal basis is for showing that warning.

In particular, it appears that this notice is sometimes displayed even in the absence of removal requests by people, the DPAs said. They asked the search engines, "Can you confirm or exclude that this is actually the case and, if so, could you elaborate on the applicable criteria?"

The A29WP committee said that new guidelines would be issued in the autumn and that additional meetings on implementing the right to be forgotten rules may be organized with other stakeholders.

Loek is Amsterdam Correspondent and covers online privacy, intellectual property, open-source and online payment issues for the IDG News Service. Follow him on Twitter at @loekessers or email tips and comments to loek_essers@idg.com


View the original article here

Non-IBM Power8 servers, chips to appear early next year

The first third-party chips and servers licensed to use IBM's Power architecture will be on the market early next year.

IBM last year started licensing the architecture so other companies could build Power servers, chips and components. The first third-party Power servers will be for cloud and high-end applications, said Ken King, general manager, OpenPower alliances at IBM's Systems and Technology Group.

Ultimately, low-end servers could use Power chips, but that's for server makers to decide, King said. Derivative Power8 chips being designed outside IBM could be used in third-party servers, King said.

IBM's Power hardware has been used in the Linux-based Watson supercomputer, which beat humans in the TV quiz show "Jeopardy." But IBM's Power server shipments have declined in recent years as buyers move to commodity hardware running on x86 chips. IBM agreed to sell its x86 server business to Lenovo for US$2.3 billion and is now focusing exclusively on the Power architecture.

The non-IBM Power servers will compete with IBM's high-end System Z and customized PureSystem offerings. But King didn't seem concerned about that, saying the reason for licensing Power to other vendors is so that the architecture will proliferate in more servers.

"It's about making Power relevant in the marketplace," King said.

Mainframes and IBM's Power are fading away, so the company had to start licensing the chip architecture, said Nathan Brookwood, principal analyst at Insight 64.

"More important for the company is to get Power out into the larger IT industry, [to] show that its got a place outside its homegrown systems," Brookwood said.

IBM last year formed the OpenPower Alliance to cooperate with other companies on hardware and software development for the Power architecture. OpenPower members include Google and Tyan, which have already shown developer boards based on the Power8 architecture. Other notable members include Samsung and Micron, which are developing memory, and Nvidia, which is developing graphics chips.

IBM recognized Power's struggles and made a smart move by opening it up to other companies, said Charles King, principal analyst at Pund-IT.

IBM may lose Power server shipments to competition, but there could be revenue from licensing, services and system deployments. Power could find some acceptance in high-performance and cloud computing. Pund-IT's King said.

"One thing Power is effective at as compared to x86 is the ability to support a larger number of virtual machines in a concurrent system. Power CPUs support classic reliability, availability and serviceability features that IBM servers are well known for," King said.

Google was perhaps intrigued by the higher level of virtual machines supported by Power compared to x86 systems, King said.

"That could translate to better VM performance and responsiveness to cloud requests," King said.

But IBM still faces an uphill battle in getting server makers to move to Power, Brookwood said.

Server infrastructure is too invested in x86 and companies will be hesitant to move to a new architecture. That requires developing software, which takes time, money and resources, Brookwood said.

"The problem with computing systems on a shrinking user and application base is they go away. It happened to DEC Alpha, Tandem NonStop, it's happened to dozens of systems," Brookwood said.

Sun Microsystems, now owned by Oracle, opened up its Sparc microarchitecture through OpenSparc, but it didn't work out. Hewlett-Packard is also moving away from the Itanium chip and providing a path to migrate to x86 chips.

But if IBM plays its cards right, there's a chance Power can live on.

"To ensure the longevity of Power8 is to get other people to use it and develop on it," Brookwood said.

Agam Shah covers PCs, tablets, servers, chips and semiconductors for IDG News Service. Follow Agam on Twitter at @agamsh. Agam's e-mail address is agam_shah@idg.com


View the original article here

Monday, July 28, 2014

Phone unlocking bill clears US House, next step is president's signature

A bill that allows consumers to unlock their cellphones for use on other carriers passed its last hurdle in Congress on Friday, opening the way for it to become law once it is signed by President Barack Obama.

Senate Bill 517 overturns a January 2013 decision by the Library of Congress that ruled the unlocking of phones by consumers fell afoul of the Digital Millennium Copyright Act (DMCA). It had previously been permitted under an exception to the anti-circumvention provisions of the DMCA, which are generally aimed at cracking of digital rights management technology.

Cellphones and smartphones are typically supplied to consumers with a software lock that restricts their use to a single wireless carrier. Removing that lock -- the process of "unlocking" the phone -- means it can be used on the networks of competing carriers. In the U.S., this is most often done with handsets that work on the AT&T or T-Mobile networks, which share a common technology, but is also popular with consumers who want to take their phones overseas and use foreign networks rather than roaming services.

The Unlocking Consumer Choice and Wireless Competition Act has made fast progress through Congress. It was passed by the Senate on July 16, just a week after it was passed by the Senate Judiciary Committee, and on Friday by unanimous vote in the House of Representatives. It now waits to be signed into law.

In addition to making the unlocking process legal under copyright law, the bill also directs the librarian of Congress to determine whether other portable devices with wireless capability, such as tablets, should be eligible for unlocking. 

"It took 19 months of activism and advocacy, but we're finally very close to consumers regaining the right to unlock the phones they've legally bought," said Sina Khanifar, who organized an online petition that kicked off the push to have the Library of Congress decision overturned. The petition attracted more than 114,000 signatures on the White House's "We The People" site.

"I'm looking forward to seeing this bill finally become law -- it's been a long road against powerful, entrenched interests -- but it's great to see citizen advocacy work," he said in a statement.

Martyn Williams covers mobile telecoms, Silicon Valley and general technology breaking news for The IDG News Service. Follow Martyn on Twitter at @martyn_williams. Martyn's e-mail address is martyn_williams@idg.com


View the original article here

Apple faces privacy suit following Chinese TV report

An iPhone user has filed a lawsuit for invasion of privacy against Apple, about a week after a Chinese state broadcaster raised security concerns about the device's location-tracking functions.

The U.S. class action lawsuit, filed by a woman named Chen Ma, alleges that Apple has "intentionally intruded" into her privacy with the iPhone's location tracking service. Apple has also disclosed the data to third parties, including the U.S. government, according to the claims.

In making the allegations, the lawsuit cites a July 11 report from the state-run China Central Television, which warned that Apple's location-tracking functions could be a security threat.

The function in question was the "Frequent Locations" feature found on iOS 7. The service records the places the user has visited, along with the duration, and is meant to provide tips, including nearby shops of interest and estimated commute times.

The CCTV report, however, claimed that the feature could be used to effectively spy on users. The data could reveal information about China's economy, and state secrets, according to one security researcher interviewed in the report.

Shortly after CCTV's investigation, Apple released its own statement, assuring users that the company does not track users' locations. Nor does it have access to the Frequent Locations function on users' phones, or has worked with any government agency to create backdoors in its products, it added.

Apple on Friday declined to comment about the class action lawsuit. The complaint was filed in the U.S. District Court for the Northern District of California, San Jose division.


View the original article here

IoT inspires new components for energy, wireless

The expected boom in demand for small, often isolated devices in the Internet of Things is driving developers to craft new types of components.

Two developments announced this week should help IoT come together. On Wednesday at a trade show in Tokyo, researchers showed off a prototype of a tiny power supply that harvests energy from vibrations in the air so remote sensors and other parts don't need batteries. And on Thursday, U.K. chip company Imagination Technologies announced a design for radio chips that can be used in small, power-sipping devices.

IoT promises to feed data from machines and sensors everywhere to systems that can deliver reports or automatically make changes in public or private infrastructure. To do that, the emerging network has to get into places that wires, cables and support teams can't easily reach, especially if those remote devices number in the millions. That makes size and power requirements stricter than ever.

One emerging technology to power standalone devices in the field is vibration harvesting, or generating electricity from movements in the environment. It can use a variety of sources, including wind, water, radio waves and human movement. For example, the motion of a person walking might help to power a small wearable device.

At the Techno-Frontier conference in Tokyo, researchers from Europe and Japan are showing off a prototype built around an electrostatic vibration harvester from Japanese automation vendor Omron with power management electronics from European research institution Holst Centre/imec. Combining the two technologies allowed them to build a vibration-based power supply that's much smaller than current models, the partners said. The DC power module they built measures 5 centimeters by 6 centimeters and may be able to shrink down to just 2 square centimeters.

The prototype is going through testing and customer input before volume production. It could be substituted for batteries in current designs or used in totally new products, the partners said.

The power supply is intended for wireless sensors for industrial applications such as equipment control and predictive maintenance. The sensors themselves could be "set and forget" devices with little need for maintenance because they wouldn't need replacement batteries. The researchers' design could be tuned for power output between 1.5 volts and 5 volts, the partners said.

Like Omron and Holst, chip architecture vendor Imagination is addressing small, wirelessly connected things. Its Ensigma Series4 "Whisper" RPU (radio processing unit) architecture extends the company's Ensigma line with a design geared to low power consumption more than high performance, said Richard Edgar, director of communications technology marketing. Rather than a chip itself, it's a blueprint for chip makers to design processor cores.

The Whisper architecture can be used for Wi-Fi, Bluetooth Classic, low-power Bluetooth Smart, NFC (near-field communications), GNSS (Global Navigation Satellite System) and other current or emerging wireless technologies. Imagination already sells the Ensigma Explorer architecture for devices such as set-top boxes that need high speed, but it designed the Whisper technology for equipment that will never send big streams of packets.

"For IoT, the criteria are different," Edgar said. "It's small amounts of data, occasionally." For example, a temperature sensor in a home might take readings every 15 minutes and send a few hundred bytes. Whisper might also be used in industrial sensors, wearable devices or smart electrical grids, he said.

Whisper is designed for use in chips built with the MIPS architecture, which Imagination acquired in 2012. It can be implemented in a MIPS chip specifically for communications or built into a single chip that also handles application processing and other tasks, Edgar said. The Whisper is scheduled to ship in a series of different versions starting in the fourth quarter.

It's hard to say whether the Omron-Holst vibration-harvesting design could power a device using Imagination's new architecture, because there are many ways to measure energy, Edgar said. But the voltages it could generate are within the right range.

Stephen Lawson covers mobile, storage and networking technologies for The IDG News Service. Follow Stephen on Twitter at @sdlawsonmedia. Stephen's e-mail address is stephen_lawson@idg.com


View the original article here