Maximizing enterprise bandwidth by sharing global enterprise-wide network resources

Here is a slightly wild idea of mine

Introduction

Generally in enterprises the network bandwidth available to employees is greatest at the start of the working day and progressively becomes bad as the day progresses. The bandwidth available improves again later in the day. The reason is that the bandwidth demand is highest during the middle portions of the day usually between 10 am – 5pm when the number of employees are the highest. The poor response times during the day are common to many enterprises as the network infrastructure is stretched beyond its designed capacity

Description

This post looks at a novel technique where the increasing bandwidth is ‘outsourced’ to other geographical regions. In other words the increasing bandwidth requirement is ‘borrowed’ from the network infrastructure from other regions.

This idea visualizes the user of a network devoice (Time Zone –Traffic Director Switch) that borrows bandwidth based on the time-of-day and the bandwidth shortage in a particular region. This device effectively borrows the idle network resources (routers, hubs, switches) from other geographical regions thus resolving the bandwidth issue.

In large global organizations, office are spread throughout the world. There will be typical daytimes and night times.  Night time usages will be minimal unless the office works on shifts

Large enterprises typically use leased lines between their various offices spread across the globe. Each of the regional offices will have its own set of routers, Ethernet switches, hubs etc.

 1

The leased lines and the network infrastructure will experience peak usage during the daytime which will subside during the region’s night times. This idea looks at a scheme in which the leased line and the network infrastructure at region B are utilized during the regions’ night times to supplement the bandwidth crunch in region A’s day time

In a typical situation we can view that the enterprise’s region as the center of the network resources (hubs, routers, switches) in its own region. Simplistically we can represent this as shown in Fig 1

 2

This post visualizes the realization of a network device namely the Time Zone-Traffic Director Switch which effectively borrows bandwidth based on the time-of-day and the bandwidth shortage in a particular region.,

This device effectively borrows the idle network resources (routes, hubs switches) from other geographical regions this resolving the bandwidth issue. Such a network device can be implemented using the Open Flow Protocol which can virtualize the network resources

In large cloud computing systems like Facebook, Google, Amazon, which have data centers spread throughout the world, traffic is directed to a data center closest to the user to minimize latency, In this scheme the use of the network resources in other geographical regions may introduce network latencies but will be economically efficient for the enterprises.

In fact this Time Zone-Traffic Director can be realized through the use of the Open Flow Protocol and Software Defined Networks (SDNs)

Detailed Description

This post utilizes a scheme in which the network resources from another are used to supplement the network resources at region during its daytimes. The leased lines and the network resources in other regions during their nights will typically experience low utilization of both the leased line capacity and the network infrastructure capacity. So in this method the bandwidth resources are borrowed from another region where the utilization of the network resources is low. This supplements the bandwidth in the region which is experiencing a bandwidth shortage. This scheme can be realized by the utilization of the Software Defined Network using the Open Flow protocol.

The Open Flow protocol has a Flow Controller which can virtualizes the network resources. This method visualizes virtualizing the network resources of a corporate office that is spread geographically around the world to optimize the corporate’s resources

This is how the scheme will have to be realized

  1. Over dimension the leased line capacity to be at least 30%-40% more than the total capacity of it Ethernet switches gigabit routers etc.  So if the network resource at a region is 16Gbps then the leased line capacity should be of the order of 20Gpbs
  2. When an external site is to be accessed for high speed connection,  after the DNS resolution the packets must be sent to the Time Zone –  Traffic Director Switch
  3. The Time Zone – Traffic Director Switch will determine the bandwidth availability at that time.  Let us assume that the regional network resources are experiencing peak usage It will not decide to use the network resources of another region
  4. The Time Zone – Traffic Director Switch will now check the time of the day.  If it is the ‘regional day’ (e.g. 10 am – 5pm, when the usages is highest) it will hard code the next hop to be a router in another geographical region. SO if the traffic is to originate in Bangalore, the next hop, if there is bandwidth crunch will be in another a router in Europe, or in the US.
  5. The Time Zone – Traffic Director Switch can be realized through th e use of the Open Flow protocol
  6. By artificially redirecting the traffic to another region all packets will be re-routed to use the leased lines & network resources from the other region thus giving the region, Bangalore, in this case, more bandwidth

This can be represented in the following diagram

 3

In the figure below The Time Zone – Traffic Director Switch uses the regions ‘internal’ network resources since it does not face a bandwidth crunch

 3

In the next figure the Time Zone – Traffic Director Switch borrows the network resource from Region B

5

This scheme will introduce additional latency as opposed to the Shorted Path First (SPF) or the Distance Vector Algorithm (DVA) However the enterprise gets to use its network infrastructure more effectively. This will result in the enterprise’s network infrastructure to be used more efficiently and effectively reducing the CAPEX and OPEX for the entire organization.

The entire scheme can be represented by the flow chart below

 6

Conclusion

The tradeoff in this scheme is between the economics of reusing the network and the leased line infrastructure of the enterprise versus the additional latency introduced

However the delay will be more than compensated by the increase in the bandwidth available to the end user. This idea will simultaneously solve the problem of the bandwidth crunch while efficiently and effectively utilizing the network resources across the entire enterprise. This scheme will also reduce the organization CAPEX and OPEX as resources are used more efficiently.

Thoughts?

Find me on Google+

A method for optimal bandwidth usage by auctioning available bandwidth using the OpenFlow protocol

Here is a recent idea of mine that has made it to IP.com (Intellectual Property.com). See link

A method for optimal bandwidth usage by auctioning available bandwidth using the OpenFlow protocol.  Here is the full article from IP.com

In this article I provide some more details to my earlier post – Towards an auction-based internet.

Abstract:
As the data that traverses the internet continues to explode exponentially the issue of a huge bandwidth crunch will be a distinct possibility in the not too distant future. This invention describes a novel technique for auctioning the available bandwidth to users based on bid price, quality of service expected and the type of traffic. The method suggested in this invention is to use the OpenFlow protocol to dynamically allocate bandwidth to users for different flows over a virtualized network infrastructure.

Introduction:
Powerful smartphones, bandwidth-hungry applications, content-rich applications, and increasing user awareness, have together resulted in a virtual explosion of mobile broadband and data usage. There are 2 key drivers behind this phenomenal growth in mobile data. One is the explosion of devices viz. smartphones, tablet PCs, e-readers, laptops with wireless access. The second is video. Over 30% of overall mobile data traffic is video streaming, which is extremely bandwidth hungry. Besides these, new technologies like the “Internet of Things” & “Smart Grids” now have millions and millions of sensors and actuators connected to the internet and contending for scarce bandwidth. In other words there is an enormous data overload happening in the networks of today.
Two key issues of today’s computing infrastructure deal with data latency and the economics of data transfer. Jim Gray (Turing award in 1998) in his paper on “Distributed Computing Economics” tells us that the economics of today’s computing depends on four factors namely computation, networking, database storage and database access. He then equates $1 as follows
One dollar equates to
= 1 $
≈ 1 GB sent over the WAN
≈ 10 Tops (tera CPU operations)
≈ 8 hours of CPU time
≈ 1 GB disk space
≈ 10 M database accesses
≈ 10 TB of disk bandwidth
≈ 10 TB of LAN bandwidth
As can be seen from above breakup, there is a disproportionate contribution by the WAN bandwidth in comparison to the others. In others words while the processing power of CPUs and the storage capacities have multiplied accompanied by dropping prices, the cost of bandwidth has been high. Moreover the available bandwidth is insufficient to handle the explosion of data traffic.
It is claimed that the “cheapest and fastest way to move a Terabyte cross country is sneakernet (i.e. the transfer of electronic information, especially computer files, by physically carrying removable media such as magnetic tape, compact discs, DVDs, USB flash drives, or external drives from one computer to another).
While there has been a tremendous advancement in CPU processing power (CPU horsepower in the range of petaflops) and enormous increases in storage capacity(of the order of petabytes) coupled with dropping prices, there has been no corresponding drop in bandwidth prices in relation to the bandwidth capacity.
It is in this context an auction-based internet makes eminent sense. An auction-based internet would be a business model in which bandwidth would be allocated to different data traffic on the internet based on dynamic bidding by different network elements. Such an approach becomes imperative while considering the economics and latencies involved in data transfer and the emergence of the promising technology known as the OpenFlow protocol. This is further elaborated below

Description

As mentioned in Jim Turing’s paper a key issue that we are going to face in the future has to do with the economics of data transfer and the associated WAN latencies
As can be seen there are 3 distinct issues with the current state of technology
1) There is an exponential increase in data traffic circling the internet. According to a Cisco report the projected increase in data traffic between 2014 and 2015 is of the order of 200 exabytes (10^18)).The internet is thus clogged due to the many bandwidth hungry applications and millions of devices that make the internet
2) WAN latencies and the economics of data transfers are two key issues of the net
3) Service Providers have not found a good way to monetize this data explosion.
Clearly bandwidth is a resource that needs to be utilized judiciously given that there are several contenders for the usage of bandwidth.
Detailed description: This invention suggests a scheme by which internet bandwidth can be auctioned between users based on their bid price, Quality of Service (QoS) required and the type of traffic (video, voice, data, streaming). The energy utility already auctions electricity to the highest bidder. This invention suggests a similar approach to auction scarce bandwidth to competing bidders.
The internet pipes get crowded at different periods of the day, during seasons and during popular sporting events. This invention suggests the need for an intelligent network to price data transfer rates differently depending on the time of the day, the type of traffic and the quality of service required. In this scheme of things the internet will be based on an auction mechanism in which different devices bid for scarce bandwidth based on the urgency, speed and quality of services required.
Such a network can be realized today provided the network and the network elements that constitute the internet implement the OpenFlow protocol.
Software Defined Networks (SDNs) is the new, path breaking innovation in which network traffic can be controlled programmatically through the use of the OpenFlow protocol. SDN is the result of pioneering effort by Stanford University and University of California, Berkeley and is based on the Open Flow Protocol and represents a paradigm shift to the way networking elements operate.

SDNs can be made to dynamically route traffic flows based on decisions in real time. The flow of data packets through the network can be controlled in a programmatic manner through the OpenFlow protocol. In order to dynamically allocate smaller or fatter pipes for different flows, it necessary for the logic in the Flow Controller to be updated dynamically based on the bid price, QoS parameters and the traffic type.

The OpenFlow protocol has a Flow Controller element which can be made to create different flows by manipulating the flow tables of the different network elements. Hence the Flow Controller depending on the bid price, the bandwidth rate and the QoS will auction the different bids and create different flows for different users. The Flow Controller will then update the flow tables of the network elements that will participate to realize this end-to-end flow of traffic for different users.

A typical scenario can be visualized as below

pat1

In the above figure different users bid for available bandwidth. For e.g. User A could bid for A Mbps @ $a/bit for traffic type A, User B could bid for B Mbps @ $b/bit for traffic type B and User C could bid for C Mbps @ $c/bit for traffic type C. The different QoS parameters like delay, throughput, and jitter are all sent in the user requests. The Flow controller receives all these bids with associated parameters and auctions the available bandwidth against the bid prices that the network elements bid for. The Flow Controller then ranks the bids against the most optimal bandwidth allocation that has the highest return.

The Flow Controller can then allocate different bandwidths to the different users based on the bids from the highest to the lowest, quality of service and the type of traffic. Software Defined Networks (SDNs) can then create different flows for across the networks. SDN can create different slices of network elements from end-to–end for each of the different flow requirements.

The Flow Controller can then create these flows and update the flow tables of the network elements based on the allotted speeds for the bid price.

This is shown diagrammatically below

pat2

 

For e.g. we could assume that a corporate has 3 different flows namely, Immediate, ASAP (As soon as possible) and  price below $x. Based on the upper ceiling for the bid price, the OpenFlow controller will allocate a flow for the immediate traffic of the corporation. For the ASAP flow, the corporate would have requested that the flow be arranged when the bid price falls between a range $a – $b. The OpenFlow Controller will ensure that it can arrange for such a flow. The last type of traffic will be allotted a default flow during non-peak hours. This will require that the OpenFlow controller be able to allocate different flows dynamically based on winning the auction process that happens in this scheme.

Using the OpenFlow paradigm to auction bandwidth

These will be the typical steps that will occur during

  1. Let us assume that it is the period of the day when the usage is at its peak
  2. Let there be 3 users User A, User B and User C who would like to video-conference, video stream and make a voice call respectively
  3. Depending on the urgency and the price that the users can afford these 3 users will bid for a slice of a bandwidth to complete their call
  4. Let user A request A Mbps @ $a/bit for QoS parameters p(a). Let user B request B Mbps @ $b/bit for QoS parameters p(b) and user C request C Mbps @ $c/bit for QoS parameters p(c).
  5. When the Flow Controller receives these requests, based on the available bandwidth at its disposal (assuming it has already used X Mbps for already existing flows) it will normalize these requests and auction them so that it results in the highest bid winning its requested bandwidth slice followed by the ones lower than it. If a user does not qualify the auction the user will have a bid at a later time according to some algorithm. Let us assume that user A and user C win their bids
  6. The Flow Controller will now algorithmically decide the contents of the flow tables of the intervening network elements and will accordingly populate these flow tables
  7. The flows for User A and User C are now in progress.
  8. The Flow Controller will accept bids whenever there is spare bandwidth that can be put up for auction.

As can be seen such a mechanism will result in a varying price for bandwidth with the highest value during peak periods and lower values during off-peak periods.

Benefits: The current protocols of the internet of today namely IntServ, DiffServ allocate pipes based on the traffic type & class which is static once allocated. This strategy enables OpenFlow to dynamically adjust the traffic flows based on the current bid price prevailing in that part of the network. Moreover the usage of OpenFlow protocol can generate a lot more granualar flow types.

The ability of the OpenFlow protocol to be able to dynamically allocate different flows will once and for all solve the problem of being able to monetize mobile and fixed line data This will be a win-win for both the Service Providers and the consumer. The Service Provider will be able to get a ROI for the infrastructure based on the traffic flowing through his network. Users can decide the type of service they are interested and choose appropriately. The consumer rather than paying a fixed access charge could have a smaller charge because of low bandwidth usage.

Conclusion: An auction-based internet is a worthwhile business model to pursue. The ability to route traffic dynamically based on an auction mechanism in the internet enables the internet infrastructure to be utilized optimally. It will serve the dual purpose of solving traffic congestion, as highest bidders will get the pipe but will also monetize data traffic based on its importance to the end user.

Find me on Google+