Google Fiber – Is it attractive? – An analysis

Google yesterday announced its Google Fiber project for high speed internet connectivity making an entry into the post decade-old broadband era.  Google has been spreading up its wings from being a mere search company. It wants to enter every segment of business which directly impact the end customers. Some say Google is still a search company since majority of the revenues still come from ads in search business, some say it is more Android focused these days, few says that Google has emerged into a hardware company after the launch of its own hardware in Nexus Q and fighting hard to convince netizens that it would also emerge as a Social Media company with tight integration of Google Plus in every aspects of its business.

I am sure Google takes a tough look at the business scenarios before making an entry move to these segments and try hard to achieve a flag bearer in ever segment of business it operates.

The latest such move is the Google Fiber project it announced in Kansas City, Missouri, USA. Google Fiber is supposed to provide 1Gbps of bandwidth as Google promised.

Here are some of the interesting points to note about:

1. Users have to pre-register in Google Fiber website for registering for a connection. The charge is 10$. This is an interesting stuff as Google wants to earn in every possible way.

2. If you are registered in Google Fiber website, it does not mean that you will get the fiber installed to your home.  Google Fiber introduced a concept called “Fiberhood” where you convince your neighbors to register for Google Fiber. Google selects the area from where most number of registrations come from.  Well, going by Google brand name, I am sure people will sign up for this (and does not really mind for the 10$ registration fee). Based on a recent study, it is determined that it costs around 650$ to lay a cable to your home. This puts a real hole in the OPEX of the ISP’s. With the Fiberhood, Google hopes  to really cut down this cost to a larger extent.

3. Google has done with laying the fiber throughout the Kansas city.  Google says that it has sought help from utilities dept to use the power poles for the fiiber and AT&T for laying the cable. Brilliant move, but strange. Why should AT&T help Google in this as the product being launched by Google is a tough competitor for AT&T. Leave me a comment if anyone knows what exactly is AT&T’s involvement in this project. Why it should help Google in the first place?

4. The pricing plans are not really competitive to encourage people to move to Google fiber. However you would get better speeds for the same rate that you are currently paying for lesser bandwidth.

5.  For folks who were thinking that Nexus Q is the only hardware device made directly from Google stable, just to let you know that Google also has a DVR device to its credit.

Leave me a comment if you have any thoughts / analysis / comments.

Have a great weekend !

Advertisements

Network Operators at the mercy of MVNO’s?

MVNO’s (Mobile network virtual operator) are entities who do not own the backbone infrastructure of the network, but instead provide the network service by leasing bandwidth’s from bigger operators. MVNO’s are entities who typically provide the service with less charge than the bigger network operators.I have always wondered as to why the network operators provide bandwidth for these so called Mobile Virtual Network Operators who eat up into the revenues and the customer base of the network operator. Recently I have read an article regarding this and here is why Network Operators seek co-operation from MVNO’s:

1)  ARPU (average revenue per user) is a critical element in judging the current financial strength of a network operator and to predict the future growth in terms of profits. In order to maintain a higher ARPU, the network operators are under pressure to maintain the high cost of the monthly plans.  Here is where the MVNO’s help the network operator. MVNO’s lease the bandwidth from the network operator. The customers of MVNO’s does not become a customer of the network operator and hence this helps the bigger network operator to maintain a healthy ARPU and hence make to maintain the momentum of its share price.

2) Since the MVNO’s lease the bandwidth from the network operator, this help to boost the bottom line of the network operator.  Apart from this , the network operators can save some more money on the overhead involved in the customer care operations for MVNO’s since MVNO’s manage their customers.

3) The network operators operating on a nation wide cannot cater to every segment of the society and hence cannot fine-tune the tariff plans for every segment of people. Here is where MVNO’s help the customer by devising plans best suited for the needs of the people. As and when the customer base increases, the MVNO’s would be willing to lease more bandwidth from the network operators and hence this will increase the revenues cum profits for the network operators from this segment of the business.

So, in short, the network operators and their MVNO’s work together for mutual benefits in terms of better revenues/profits as well as better care for the customers who cannot be served by the bigger network operators.

Software Defined Networking – The next big networking wave !

Earlier I have blogged on CDN / TIC where the online content is being moved from the servers in content providers network to the networking gears in the service providers network in an aim to achieve higher speeds for content access. In contrary to that, SDN (Software defined networking) specifies a mechanism where the content could still be present in the content provider network, but the users could still achieve higher speeds in terms of accessing the contents if the gears in the network is fine-tuned.

Earlier this week, Google announced about its super secret Networking project powering its inter-data center  communication using SDN for most (all?) of Google’s services – Search, GMail etc. In addition to the brilliant algorithm that Google implements to retrieve the search results quickly, SDN also plays a major role for those magic speeds. Google calls this network as “G-Scale Network” which replaces the traditional routers from the networking giants with routers running Open source softwareOpenFlow.  It take real guts to run the network backbone infrastructure using experimental protocol like OpenFlow.  To take part in the growth of SDN, several OEM (original equipment manufacturers) started introducing the products which can run OpenFlow. The individual companies buying such products can opt for implementing the OpenFlow protocol.

How does OpenFlow work / operate?

Typically in a OpenFlow environment, the control plane will be running on independent servers and the data plane would be   running on the routers. The communication between the control plane and the data plane would be using the OpenFlow protocol.   The servers powering the control plane and equipped with better processors and the software to provide a better performance  when compared to the processors running on the network routers.  In addition to the performance improvement on the control  plane, it also provides better security for the control plane.   Ofcourse, this is at a very early stage and takes time to stabilize. 

Cisco which has been known for making proprietary software for routers have started facing the heat with the introduction of SDN  – which is going to define the networking in the decade to come. HP, IBM, Nicira Netowks, Big Switch Networks have already intoduced products supporting OpenFlow for SDN. To make sure Cisco also participates in the race for the decade defining innovation and does not loss the market share in SDN, Cisco  is planning to introduce SDN hardware in the next 6 months to a year.

Here are some of the key points about the notification from Google during the OpenNetworking summit .

The official agenda for OpenNetworkingSummit has some interesting information. I liked the first few slides from “Big Networks”  from Open Networking Summit 2012.

In my opinion, CDN and SDN would not be competing to win the title for “War of Titans”, but instead they would/should work together for enabling a powerful networking environment.

Also I have a workshop on SDN in mid of May and will be updating the blog with more information as I learn.

Please leave a comment on the question/thoughts.

Content Delivery Network – Primer

Content Delivery Network or Content Distribution Network is a group of network servers located in the data centers around the world aiding in the faster retrieval of  contents to the internet users.  These distributed servers cache the static information for a content providers. CDN’s help to retrieve the static contents upon user request from a geographically closer CDN server using an algorithm, thus enabling a quicker response for the users request. Without CDN, these requests would be delayed especially if the client and the content provider are far apart. The dynamic content would still be available in the content providers network. The static contents that are being provided have some sort of versioning in the contents. If the appropriate version of the static content is not present in the CDN servers, then a cache miss occurs. Whenever there is a cache miss, the CDN servers request for the static data from the content provider server and cache the data in their local servers. By this way. any request for accessing the static content can be full-filled by the CDN servers.

CDN’s either use their own data centers or use the data centers from ISP. The content providers pay the CDN provider for ensuring the reliability and efficient access to the contents. The CDN providers inturn pay to the ISP for hosting the servers in the data centers if they use the ISP data centers.  These geographically located CDN servers are called surrogate servers.

Akamai is one of the most popular and early entrant in the CDN world.  In the CDN world, the servers being managed by the CDN are called surrogate servers. These surrogate servers are interconnected to each other. They also form a hierarchical patters – parent <-> child surrogate servers. These surrogate servers in the network are typically connected by  a “CDN pairing gateway” which interconnects two or more CDN which are separately administered.

These CDN’s also interact with the accounting system of the content provider so that the contents are provided to the users based on the subscription validity of the clients.

The core and access routers in the network are powered to pull the data from the appropriate surrogate servers (or these routers act as surrogate CDN servers caching the information?) thereby enhance the speed of retrieval of data.  The standards for the CDN network are not clearly defined as of now and would certainly empowered the next generation of the internet boom (esp for video services).

Google – From “don’t be evil” to “try to be evil”

Image representing Google as depicted in Crunc...

Google is requesting Microsoft to pay a hefty fee for use of its 50 H.264 video standard patent which it acquired from Motorola. These patents are used for accessing video services over Wifi. Google is claiming Microsoft to pay 22.50 USD for every Microsoft Windows  laptop less than or equal to 1000 USD. The patent fee doubles for laptop that are 2000 USD worth. This price is ridiculous when compared to the amount that Microsoft currently pays for use of 2000 odd licenses – which is less than 2 cents to 29 companies. Not sure if this fee would be for the video access over Wifi for the Windows mobile and tablets.

Image representing Microsoft as depicted in Cr...

Microsoft is ready to pay for a fee that is justifiable and not ready to pay such a huge amount. I think the huge fee is an attempt for the patent fee that Microsoft charges Google for every android phone activation. Microsoft has filed a compliant with the European Commission and the US. Department of Justice seeking investigation the patent fee.

Here is a report from Dave Heiner, Corporate Standards, Microsft. Snippet of the blog post is as below:

<Snip>

There is an obvious way out of all this. Motorola should honor its promises, and make its standard essential patents available on fair, reasonable and nondiscriminatory (FRAND) terms. Microsoft is certainly prepared to pay a fair and reasonable price for use of others’ intellectual property. Within just the past few years, Microsoft has entered into more than a thousand patent licenses. We know how it’s done.

</Snip>

Is Cisco trying to secure the future of Video communication?

Image representing Cisco as depicted in CrunchBase

Cisco, a leading networking gear conglomerate has appealed    before the European Union concerning the fact that the merger of Microsoft and Skype would create a monopoly in the Video communication space. With the 8.5 billion dollar buy-out of Skype by Microsoft, Cisco feels that Microsoft would be able to leverage on the 700 Million + user base of Skype in its Lync platform and would be attempting to create a monopoly where its service would not be compatible with the services from other vendors. MessageNet, an European VOIP provider have also joined the race.

Cisco believes in the future of the Video communication and prepares itself as a networking vendor as well as a provider of its service platform “Telepresence“. In the recent years Cisco has introduced high powered routers for handling terabytes of video traffic.

Although Cisco does not really oppose the acquisition of Skype by Microsoft, all it wants is to create a Open standards based Video Communication ecosystem where users in multiple platform can inter operate.  Cisco’s concern is pretty fair. The next question which needs answer is what Cisco going to gain from this battle? I do not agree to the fact that Cisco is fighting for this, without a clear strategy for a “win” situation. Would the empowerment of user in adapting the open standards based architecture increase the network traffic and with this Cisco expects more sales in its high-powered routers? Could be.

In my opinion,

– Video communication should use open standard based architecture which would enhance innovation in the network and branch out services / applications based out of Video.

– Users of the video service would be greatly empowered with an inter-operable platform.

Hopefully European Union Commission would revisit the deal approval and request Microsoft to adhere to Open inter-operable based standards.

Here is a blog release from Marthin De Beer, SVP, Emerging Business Group from Cisco