Feb 2, 2011

Opening the Door to In-Building Copper Network Upgrades

Many communication services providers have made the decision to deploy fiber-to-the-building (FTTB) networks to multi-dwelling-units (MDUs). This hybrid approach, in which the existing in-building copper network remains active, can be a cost-effective and viable approach for delivering the high-speed services subscribers demand while keeping costs in check.

Yet, providers must ask themselves: Will this strategy work 5 years from now? What about 10 or 20 years down the road?

Enhanced services are in their infancy in terms of adoption and development. Social networking, cloud computing, high-speed file transfers, and streaming audio and video will continue to change the way bandwidth is used.

Copper has so far proven its ability to support high-speed services. Many providers, both large and small, have successfully deployed broadband over Last Mile copper networks. However, providers must prepare those networks for the future.

Below are 6 steps to upgrade your in-building copper networks to meet the ever-increasing demand for enhanced services.

Step 1. Improve performance and density of in-building entrance terminals (IBETs).

Network reliability is more important than ever because degradation of the network is more noticeable than ever. Higher bandwidth services operate at higher frequency electrical signals in the copper network. These higher frequency signals are more prone to imperfections in the legacy network, which can manifest in higher crosstalk, higher signal loss, and unbalance. Today, with multiple channels of HD video and high-speed Internet, and higher subscriber expectations, those imperfections are unacceptable. Pixelation or delays during a favorite TV show leads to angry customers.

The connectivity or cross-connect solution chosen for the in-building entrance terminal (IBET) directly affects the quality of service to subscribers. Contact integrity provided by the connector is critical to ensure reliable service, not only at the time of installation, but over the life of the connector.

Replacing a building’s aging 66 blocks with VDSL-compatible and Category 5 (CAT5)-compliant blocks can help achieve network reliability. Choosing a gel-filled, easy-to-install, insulation displacement contact (IDC) block, goes even further to protect reliability. Gel encapsulation helps protect the connection against moisture and corrosion, preserving a high-integrity, reliable connection for long-term broadband performance. The ideal IDC block could be installed without stripping wires or using tools, helping reduce the risk of it being installed incorrectly. It would also be able to be re-opened for visual validation of the IDC connection.

Where space is a concern, consider equipment that will increase density in the IBET. A typical indoor building entrance terminal today is configured with the terminal blocks installed along with separate 5-pin protective devices that shield equipment from transient voltage and current surges. Copper cables connect the individual subscriber blocks with the protection devices.

Recently, 3M introduced an innovative terminal block design (3M™ Quick Connect System 2814) that integrates the protection into the block. (It is currently deployed in other regions of the world and is anticipated to be release in the U.S. later this year.) (See Figure 1.) With the block, the entrance terminal holds just 1 device per subscriber instead of 2, essentially doubling density. Now, a 50-pair box can hold 100 pairs, allowing for more subscribers or additional services. The dual-purpose block also supports better signal integrity by eliminating the connection point between the terminal block and the protection device, where signal can degrade, potentially interfering with service.

Step 2. Optimize POTS splitter management for better service quality and greater efficiencies.

FTTB/VDSL2 applications with a remote DSLAM at the premises commonly require the installation of 2 sets of DSL-dedicated terminal blocks and equipment cables to and from the DSLAM at the building’s main distribution frame (MDF). To provide DSL service to a customer, 2 jumpers -- POTS IN and DSL OUT -- will run between the existing MDF and the DSL terminal blocks (See Figure 1.)


Figure 1. FTTB/VDSL2 applications with a remote DSLAM at the premises commonly require the installation of 2 sets of DSL-dedicated terminal blocks.

For larger, non-”pizza box” remote DSLAM applications (typically 50 or more subscribers), it is possible to achieve a higher density infrastructure with lower installation costs by using a VDSL2-ready block with integrated POTS splitters with just 1 equipment cable between the DSLAM and the MDF.

Besides creating such a high-density system, a well-designed integrated splitter block can also help deliver enhanced VDSL2 service. Individual wire guide channels can help maintain the wire pairs’ tight twist up to the base of the block’s Insulation Displacement Contacts (IDC). Sustaining the twist throughout the cable length reduces the potential for crosstalk and supports VDSL signal integrity. POTS splitters mounting on the block’s rear side will reduce the chance of accidental disconnections.

VDSL2 POTS splitters need to be fully compliant with the requirements of ANSI T1.413-1998 Issue 2 Annex E, ITU-T G.993.2 and designed to include robust ring trip protection with full compliance to the Broadband Forum TR 127 Issue 2 to support quality multimedia transmission.

Step 3. Provision a DSLAM port to every living unit.

For larger deployments where a high subscribership rate is expected, pre-deploying a DSLAM port to every living unit can reduce installation costs down the road. By using a VDSL2-ready block with integrated POTS splitters, a DSLAM port can be assigned up front to each living unit. To activate DSL, the technician skips the step of removing the POTS jumper and installing 2 new jumpers because it’s already done. The technician simply connects the already-installed jumpers to the splitter block.

Providers have shied away from taking this tact, and rightly so. Assigning each living unit with a DSLAM port requires a big initial investment in materials and labor. However, in buildings where DSL service is typically high, such as high-end apartments or college dorms, pre-deployment helps make activation easier and is cost-effective over time.

Step 4. Upgrade your copper cable to at least CAT5e.

The conduit that carries signal from the building entrance terminal to the individual subscriber is obviously an essential part of the in-building network.

Category 3 (CAT3) cable is primarily used for POTS wiring and still widely deployed today. CAT3 cable can support low-speed data applications with frequency transmissions up to 16 MHz and data transfer rates up to 10 Mbps. That simply won’t cut it for high-speed services available today and in the future. When deploying DSL, CAT3 should be replaced with a higher performing cable (i.e., CAT5 Enhanced [CAT5e] or better).

CAT5e cable is specified for transmission frequencies up to 100 MHz and can support data rates up to 1 Gbps, providing optimal performance for all data and phone systems.

CAT6 marks the next step up from CAT5e and carries a higher price tag. CAT6 cable raised the bar for operating frequencies to 250 MHz and can support data rates up to 10 Gbps.

The newest cable specification, CAT7 cable, is now also available, although it has not yet been widely adopted except for some government and manufacturing applications. It can achieve 600 MHz and 10 Gbps.

The difference is in the twist. Generally speaking, the tighter the wire pairs within the cable are twisted together, the better the signal integrity. When wires are laid side-by-side, the electromagnetic signal traveling through can jump from one wire to the other. The resulting crosstalk interferes with the signal, causing data errors, service interruptions and unhappy subscribers. At low frequencies, crosstalk usually doesn’t cause many problems. But at high frequencies, crosstalk can degrade the signal enough to interfere with service. Twisting the wires together has a canceling effect on crosstalk. The more twists per inch of wire, the better the cancellation effect.

For instance, CAT3 cable wires are twisted 3 to 4 times per foot of cable. CAT5e cable contains 3 to 4 twists per inch of cable, 12 times the twist of CAT3, leading to significantly better signal integrity (and happier subscribers).

Step 5. Train technicians to inspect for proper bonding and grounding.

As building communications closets become crowded with more and more equipment, proper bonding and grounding becomes increasingly important to ensure safety of people, property and the network. Proper bonding and grounding ensures that foreign voltage safely drains to the earth ground. Poor bonding and grounding can lead to service outages, damaged equipment, and, worst of all, the possible injury or death of a person.

Most operators have their own procedures for bonding and grounding based on industry guidelines. Still, due to lack of training and oversight, it is not uncommon to find improperly grounded telecom equipment in the field. To avoid dangerous situations, communications providers should properly train technicians on proper bonding and grounding, and continually stress its importance.

Before installing service, technicians should make sure that the building developer has provided a proper ground source. The utility/communications room should be equipped with a copper-clad, fire-rated multi-grounded neutral (MGN) conductor source that is #6 American Wire Gauge (AWG) or larger, and in good condition. The ground wire should be installed as straight as possible or make sweeping bends. (See Figure 2.) Tight bends or kinks increase the risk of foreign voltage jumping off the conductor and energizing surrounding metal. If the ground wire does not meet these minimum requirements, technicians should report the situation to their supervisor and delay installation until the grounding components are brought up to standard.


Figure 2. Ground wire should be installed as straight as possible and ensure the building developer has provided a proper ground source.

The MGN ground wire source should be connected to a bus bar. Telecom equipment on the wall is also connected to the bus bar to create a path to ground. If the bus bar is full, the technician should install an additional bar (or ask the building manager to do so) and strap the 2 bus bars together. The bonding conductor for passive equipment should be jacketed, fire-rated #6 AWG or larger. Active equipment should be bonded using, at minimum, #2 AWG tinned copper conductor. Again, bends should be sweeping and not tight or kinked.

Of note, some technicians may believe that because the building terminal cabinet contains protection coils, it is safe from foreign voltage regardless of the ground wire. However, if the building terminal is not properly bonded to ground, the metal cabinet protection may prove ineffective.

Step 6. Ensure a multiple-free network.

In-building copper networks should be multiple-free to ensure high-quality service now and in the future. That means each subscriber has a dedicated pair leading from the cross-connect cabinet to the building terminal to the living unit.

Many older networks were built in a bridge-tap configuration in which multiple living units are connected to a single pair running all the way from the cross-connect cabinet. Cable counts in multiple make it easier and less costly to add new subscribers, and it may have worked just fine for POTS; however, it doesn’t work for broadband. Each tap siphons off signal strength, leaving subscribers downstream with degraded service. Imagine a garden hose with holes punctured in it every few feet. By the time you get to the end of the hose, not much water is coming out. That’s essentially how it works with bridge tapping.

New networks should be designed multiple-free and old systems reconfigured to prevent bridge tapping. One possible way to ensure a multiple-free network is to choose a cross-connect block that does not allow bridge tapping. - source

Jan 7, 2011

DIY Flash EEPROM BIOS Programmer

I got this Gateway Dx4300 computer motherboard with a dead BIOS chip in it, that wont start at all, so I figured it must be a corrupt bios. The motherboard was kinda nice that supports Phenom II quad processors & DDR3 ram (I think) .. so I didn't want to waste it just because it had a bad bios chip.


Buy a replacement bios chip you say ? .. sure , but i dont feel like spending $25-$35 right now for a motherboard i got for free .. so I had an idea to build a simple BIOS programmer that I can build & use to save his board and possibly others in the future ..

so the bios chip was an Macronix MX25L8005 (1MB) .. and .. luckily, it sat on a socket (not soldered in) ..


After some research, I found a few DIY kits & projects along with comercial ones like the Willem, Acron, SMARTPRO ..etc , they were all expensive. Eventually I stumbled upon a design that will probably do the job, so I built the simple circuit & started experiminteng with it .. a few attempts later, I was able to reflash the chip with a new bios that saved the motherboard from the garbage man, later I decided to add some improvements to this design to make it more convenient to use in the future.

My design is based on RayeR's with slight improvements, I have added a voltage regulator to power it directly from USB port, a switch and some LED indicators ..

Supported chips (and possibly others):

AMIC
SST
ST Microelectronic
A25L05PU/PT (64kB)
A25L10PU/PT (128kB)
A25L20PU/PT (256kB)
A25L40PU/PT (512kB)
A25L80PU/PT (1MB)
A25L16PU/PT (2MB)
A25L32PU/PT (4MB)
A25L64PU/PT (8MB)
A25L512 (64kB)
A25L010 (128kB)
A25L020 (256kB)
A25L040 (512kB)
A25L080 (1MB)
SST25VF010 (128kB)
SST25VF020 (256kB)
SST25VF040 (512kB)
SST25VF080 (1MB)
SST25VF016 (2MB)
SST25VF032 (4MB)
SST25VF064 (8MB)
SST25VF128 (16MB)
SST26VF016 (2MB)
SST26VF032 (4MB)
SST26VF064 (8MB)
M25P10 (128kB)
M25P20 (256kB)
M25P40 (512kB)
M25P80 (1MB)
M25P16 (2MB)
M25P32 (4MB)
M25P64 (8MB)
M25P128 (16MB)
EON
Winbond
Macronix
EN25P16 (2MB)
EN25P32 (4MB)
EN25P64 (8MB)
EN25P128 (16MB)
EN25F16 (2MB)
EN25F32 (4MB)
EN25F64 (8MB)
EN25F128 (16MB)
W25X10 (128kB)
W25X20 (256kB)
W25X40 (512kB)
W25X80 (1MB)
W25X16 (2MB)
W25X32 (4MB)
W25X64 (8MB)
MX25L1005 (128kB)
MX25L2005 (256kB)
MX25L4005 (512kB)
MX25L8005 (1MB)
MX25L1605 (2MB)
MX25L3205 (4MB)
MX25L6405 (8MB)

Spansion
Atmel

S25FL004A (512kB)
S25FL008A (1MB)
S25FL016A (2MB)
S25FL032A (4MB)
S25FL064A (8MB)
S25FL128P (16MB)


AT26DF041 (512kB)
AT26DF081A (1MB)
AT26DF161A (2MB)
AT26DF321 (4MB)




NOT BAD AT ALL FOR A $5 HOME-MADE FLASHER !!!


The finished PCB in action, just remember to erase the chip first before re-programming, it works best that way... the extra 3 pin white-connector you see is for my personal use to get +5VDC & +3.3VDC directly out (optional) .. you don't need to have that.

Software ???

There are many option out there that work with SPI based LPT port programmer .. I tested a few, SPIPGM (DOS) worked best for me .. the pic below is from a Window$ based GUI that I also tested .. and finaly, I will definately make a nice housing for this PCB and add a ZIF socket to make it more convenient to use ..


And off-course .. here's the schematic ..


This project being written by Nader Gator all works are credited to him.

Aug 1, 2010

Ceragon Equipment Powers Digitel’s Expansion

Digitel Mobile Philippines will invest $9 million in Ceragon Network’s mobile backhaul equipment. It will help with the ongoing expansion of Digitel’s 2G/3G cell services. Sun Cellular, the operating name of Digitel is the fastest growing teleco in the Philippines.


The expansion project will provide 3G services in the southern part of the country. It will use native IP over long-haul microwave technology. The other expansion, in the northern part, will provide more 2G capacity.

Ceragon equipment will help Digitel migrate to an all-IP Ethernet backhaul network. It will provide better range of data, multimedia applications and voice on the operator’s expanded network. Ceragon has a proven track record of highly available, robust, cost-effective turnkey solutions and services that made it simple for Digitel to choose the equipment provider.
“Microwave is an ideal approach to quickly and easily setting up the most cost-efficient and consistently reliable backhaul network,” Patrick Lam, Senior Technical Adviser, of Digitel Mobile Philippines Inc., said “Our Ceragon-based backhaul even withstood last year’s devastating typhoon and flooding, making Digitel the only operator able to keep delivering vital telecom services to people in some of the devastated areas of the country.”
Digitel will use a mix of TD and IP-based systems from Ceragon. The TDM-based FibeAir 1500R, high-capacity SDH/SONET wireless system and the FibeAir IP-10 which combines IP and TDM networking for migration to IP.

The FibeAir IP-10 is Ceragon’s next-generation carrier grade wireless Ethernet solution. It combines IP and TDM networking, enabling mobile backhaul migration to IP with the high capacity at low cost. The FibeAir 3200T provides increased capacity, long-distance linking and simplicity of installation.

The FibeAir 1500R is a high capacity SDH/SONET wireless system that answers the need for future-proof efficient wireless solution. The system includes a variety of interfaces that get seamlessly integrated in many different network applications. It works at long, medium and short transmission ranges with different protection schemes.

Mar 27, 2010

Programmer SPI FlashROM for parallel port

March 23, 2008 Since I'm such a meddler, as drilled in codes like BIOS, and he will not rest until the screen for good nezčerná :), and also need to be able to manually reprogram FlashROM. I used it in parallel flashek that were on the board at DIL32 or PLCC32 socket, solved simply hotflash when I had prepared in advance of the second chip with a working BIOS. The one I put into the slot at runtime exchange for badly programmed chip and reprogrammed.


Now I'm on my new motherboard Gigabyte GA-P31-DS3L met with a new type of memory - a high speed SPI FlashROM Winbond 25X40VSIG SMD SOIC-8 housing, which is soldered directly to the board. South bridge intel ICHx longer time support both variants FWH and SPI. Using certain chipset registers and pins can select from the interface is primarily used for BIOS. 

Due to the fact that with SMD components save a dime at base, overall reducing the costs of production and thus managers have more $ for better cars and bitches ;). Šťouralům then we have no choice but to take the hand solder ... To flash with that reasonably could work, I soldered it to the precise DIL8 slot. The second slot DIL8 I then soldered to the motherboard. Because I would be on-site with a solder around the base (opposite the SATA connector) did not make it, I let it dangle on two bits of 4-core braces:

Followed by a production programmer. SPI devices to connect to the PC but just an ordinary parallel port, 5 wire, a few resistors and connectors. Involvement based on the type of cable programming SPI BSD programming Atmel AVR microcontrollers. Advertised W25X40V memory is designed for 3.3 V power supply, so I separated the data line resistors, to suffer from a higher voltage LPT. At the same time acts as a damping resistors, otherwise it will on the long (about 1.5 meters) flat cable without shielding between the signal lines causing crosstalk and thus incorrect data transmission. Interestingly, the same AVR MCU me a lead they work without problems. I used to power an external source.


The whole thing then jumbled the universal plošňáku looks something like this:


Another thing that is only software that I had to write. First, it must be programmed lowest layer that works with LPT registers and sets / reads the required level of the SPI lines. On this topic, I refer to the book Burkhard Kainka - Using the PC port of Publishing HEL 1997, from which I once learned how to blink his first light on LPT. But the Internet is a resource on this topic more than enough.

In addition, I said something to the SPI bus itself. It consists of three lines: SCK - Serial Clock (up to tens of MHz), MISO - Master Output Slave Input, MOSI - Master Output Slave Input and in this case there is also an auxiliary control unit # CS - Chip Select. Communication typically takes place between two devices, one of which is the master (in this case the PC), which controls the SCK clock (there is a resting level log. 0) and sends commands / data on MOSI line while simultaneously reads the data from the MISO line. 

Slave devices (in this case, memory) then responds to the SCK clock by the falling edge of the clock sends a bit on the MISO line and the rising edge of the bit clock reads MOSI lines. Byte is sent in MSB first, LSB last. After the transfer is complete master clock returns to the idle level (logical 0). The bus is to be opposed by I2C full duplex. Slave devices do not identify any address. In order to connect to multiple lines SPI slave devices must be master with any device connected to a separate control line CS # to log levels. N 0, activates one slave device (other SPI outputs are in high impedance state to affect the ongoing communication (SPI devices have 3-state outputs are required pull-ups). In the case of SPI Flash memory, CS # is used for the beginning and end of the frame so it can not be easily connected to the ground. following picture tells more as being read from memory:


First Master starts moving beyond CS # to log. 0 and thereby activates the memory. Then starts the master clock and transmits the first byte command 3h - reading from memory. Memory determines, what we want to her and knows he has yet to address. Then the master sends the next 3 bytes. Memory address decoding and immediately starts the next byte to send data. If Master requests data from the following addresses, lets start the clock and automatically increments the memory address. Further incoming data from the Master are ignored. End of the frame, then the Master terminates transfer hours in standby log. 0 and CS # going to idle status log.

First memory supports a standard set of commands, which include the need to read, write, delete pages, erasing sectors, deleting the whole chip, lock, unlock, identifying the manufacturer and type of chip JEDEC, reading the status register, ... Some commands are byte, others have parameters and data. Some memory corresponds sending data to another Sun For more details see datasheet for specific memory.

Further step was therefore write functions for sending and receiving bytes and data blocks, and finally processing functions commands of Flash memory. The current version of the program allows you to identify the type of memory / i, to read and display the data block of a given size from a given address / r, read the entire file into memory / d, the entire program memory from / p, delete the entire memory / ea unlock write protect bits / u . Other parameters / L = Adjustable base address of LPT / d = extension of SCK pulse in microseconds (except Windows version, where there is no timer with a higher resolution than 1 ms). The download package are 3 versions for DOS, Windows and Linux.

Due to the limited speed of LPT and serial transmission method takes a read or write tens of seconds to minutes. Chip erase time depends on the speed of the internal mechanisms in memory and takes about units to tens of seconds. Faster communication could be achieved using microcontroller, preferably with a hardware SPI controller, which communicates with the PC either 8-bit after LPT or USB. Sample event program memory M25P32 Wed: More information at Rayer's Blog.

Feb 11, 2010

WiMAX enjoys its day in the sun


New data from TeleGeography’s 4G Research Service show that there were more than 600 WiMAX networks either live or at the planning/deployment stage by the end of 2009. The number of WiMAX deployments significantly exceeds HSPA and LTE deployments. Only around 300 cellular networks have been upgraded with HSPA technology and fewer than 70 operators have committed to deploying LTE equipment.

However, only a relatively small number of these WiMAX systems offer wide coverage areas; most only offer local or regional service. Late 2009 saw the introduction of the world’s first commercial LTE networks, with TeliaSonera beginning to offer limited services in Stockholm and Oslo, and 2010 is expected to bring further service launches in Europe, the US and Asia. TeleGeography analyst Peter Bell comments, "While the footprint of WiMAX worldwide is still growing steadily, LTE deployments should gain momentum in 2010 and 2011. With LTE receiving strong backing from major cellular operators and equipment vendors, we project that LTE subscribers will ramp up far more quickly than WiMAX subscribers."

Nevertheless, WiMAX will continue to grow. There is still significant demand for wireless broadband access in developing countries and in rural areas of developed markets. Moreover, the next incarnation of WiMAX – 802.16m – is already on the horizon and promises to arrive well ahead of the next generation of LTE. Said Bell, "The answer to the question ’LTE or WiMAX?’ is both. LTE and WiMAX both have roles to play, and both technologies will be with us for years to come." - source

Feb 1, 2010

Apples Lifting The Ban On VoIP Over 3G


A new iPhone developer SDK is bringing us VoIP over cellular networks. Few iPhone applications have that feature built in, but will the iPad get it as well?

Apple Lifts 3G VoIP Restrictions, iCall with 3G Support Available Immediately

iCall VoIP services for the Apple iPhone and iPod Touch are now available for use over 3G networks such as AT&T wireless.

GREENWICH, Conn., Jan. 27 /PRNewswire/ — Apple Computer, Inc. has updated the iPhone developer SDK to allow VoIP over cellular networks. iCall is the first and only VoIP application that functions on the iPhone and iPod Touch over cellular 3G networks.

iCall is a market leader in desktop and mobile-based desktop calling platforms, including its highly popular version for the Apple iPhone. Until today, restrictions imposed on developers prohibited VoIP functionality on any 3G network.

VoIP on the iPhone has been the source of many debates with companies such as Google being denied access to the platform and the FCC inquiring as to the restrictions in the Apple App Store; iCall is one of the few VoIP applications that has been permitted by Apple to operate on the iPhone platform.

With the latest revisions Apple has made to the iPhone developer agreement and Software Development Kit, iCall for the iPhone and iPod Touch now enabled unrestricted free local and long distance calling over 3G data networks. iCall with support for VoIP over 3G networks is now available in the App Store for download. iCall is the first and only VoIP application available for the iPhone platform that allows use over 3G networks. iCall for the iPhone and iPod Touch may now be downloaded from the App Store here.

iCall CEO Arlo Gilbert is quoted as saying, "I applaud Apple's decision to allow iCall to extend its functionality beyond Wi-Fi and onto the 3G networks. This heralds a new era for VoIP applications on mobile platforms, especially for iCall and our free calling model. I hope that now more developers will begin using our VoIP as a platform to integrate VoIP into their applications."

About iCall, Inc.:
iCall is a privately held company providing a free and low-cost VoIP calling platform that is currently available for your desktop PC, iPhone and iPod Touch. iCall saves consumers money on phone calling through ad-supported calling.

iCall was founded in 2005 by Arlo Gilbert and Andy Muldowney with the goal of providing low and no-cost calling services to consumers. In January 2006, iCall first released its iCall Free Calling desktop application. iCall's free calling network now boasts over more than 4 million unique downloads.

In addition to consumer products, iCall offers wholesale VoIP solutions and a developer platform through iCall Carrier Services. iCall's private network carries over 350 million minutes of voice traffic per month to destinations around the globe. - source

Jan 1, 2010

OpenDNS vs Google DNS Performance Comparison

Google recently launched it’s own Public DNS offering, allowing Internet users to make use of their DNS servers for resolving address queries on the Internet. I have been a fan and user of OpenDNS for quite some time now, which offers the same service with many more options (as compared to Google’s new service). Immediately, I was curious as to how Google’s service compares with OpenDNS.

On a functional and service offering standpoint, I don’t need to really talk about anything, because the founder of OpenDNS already has excellent thoughts on this topic. But he doesn’t talk about the most important metric of all, which is performance. I was curious about how the two compare on pure performance terms. So, sitting in India, I decided to run a quick test myself to measure the performance between the two.

The method is the same as I used when I compared OpenDNS with Airtel, my local ISP in India. I basically measured lookup times of 100 random domains and compared the two DNS on that metric. Here are my findings.

The graph below plots the query response time for both the DNS services for 100 random domains. What we can see from the graph already is that neither one is clearly faster than the other.


Here’s another graph which gives clarity on which one is faster more often:


As we can see, they are almost the same. So, how do we know which is better or faster?

I took a look at the average of the measurements for the two services, and this is what I found:
  • OpenDNS – 0.38
  • Google Public DNS – 0.41
This means that on average, OpenDNS is slightly faster than Google. However, as we can see there are 3 spikes in the Google Data, and one in the OpenDNS data. When I remove those, the average for the two is almost the same (with OpenDNS being marginally slower).

Another thing worth noting however is the Standard Deviation – the measure which tells us how often will the measurement fall close to the average performance. This can even be surmised by the first graph. With the complete measurements, the Standard Deviation of Google was about 3 times that of OpenDNS, and even with the spikes removes, the standard deviation of Google was twice that of OpenDNS. Which means that OpenDNS on the whole is more consistent in its operation.

This means that when Google is deviating on the faster side, it will be much faster than OpenDNS, but when it deviates on the slower side, it will be much slower than OpenDNS (on average).

At this stage, I will have to say that I cannot conclude, though I would like to think that OpenDNS is a better performer for providing consistent performance, as compared to Google Public DNS. But with Google’s muscle, I am sure that it will eventually surpass OpenDNS on these parameters. Who knows, maybe Google will buy OpenDNS (although I am not sure if they will be selling). - source

Jan 1, 2009

Making VoIP calls with iCall from iPhone

I’ll admit that I don’t make a ton of calls on my mobile phone, and generally don’t worry about how many minutes I’m using. However, there are those out there that talk considerably more than I. For those people I give you iCall.


iCall is a useful application which will allow you to instantly switch between VoIP and GSM calls. This means you can sit at home and use your local Wi-Fi network to make calls without using up valuable minutes.


No, the iPhone won’t be the first phone to take advantage of this sort of feature, but you can bet that it will be one of the more useful apps. This is said to already have passed the Apple inspection, and should hopefully appear on the App Store sometime after launch. - source

Jan 7, 2008

Paypal cancel automatic payment


Paypal recurring payments allow seller to automatically renew the subscription payment on the expiry date. In other words, if you bought the subscription using Paypal, your subscription will be automatically renewed and your debit/credit card will be charged at the time the billing cycle is triggered. This process continues until you cancel the service.

To stop recurring billing charges:

1) Login to your Paypal Account

2) Put your mouse over the “History” and click “Basic Search” as shown below.

You’ll get a list of all your subscriptions and their status as shown below.

3) Now click on “Details” or the name of subscription you wish to cancel.

4) At the Subscription page, Click “Cancel Subscription” as shown below.

5) On the next page, you will be asked to confirm the cancellation. Click the “Cancel Subscription” button to complete the cancellation.

6) You are done!

Jan 1, 2007

Krone 10pairs PCB Connector