If Broadband Was Free And Instant .... Or What COULD Be The Future Of Broadband In A Virtual World

This is obviously a hypothetical question ..... and I encourage you to brainstorm and leave your own ideas as comments.

Now here goes .......

If WAN connections cost practically nothing ... and were so fast and reliable that errors, bandwidth, capacity, and latency were never an issue .... then what amazing revolutionary uses can you think of?

Here's a few to chew on and get the juices flowing ...

1) Appropriate use of location beacons to aid traffic flow (appropriate = without compromising personal privacy) and therefore optimise fuel consumption and speed travel times.

(2) Background health monitoring - e.g. for those with chronic diseases such that the growing array of health-tracking sensors could be something that facilitated less stressful travel outside of the home environment.

(3) Security enabled private monitoring - again allowing for privacy considerations, it would be great if as parents we could glance at our mobile phone and know where our teenage kids were at all times. Naturally you would only want the tracking to be available to authorised users and not subject to abuse.

(4) Genuine energy consumption optimization management within office buildings and private homes - smart AI based programs could monitor building occupancy levels and automatically turn on and off lights. Smart management of air conditioning and heating. Kicking in low energy standby mode for a whole host of appliances.

For me there would be a whole host of quality-of-life enhancements that would become practical and indeed cost-effective with a ubiquitous broadband "cloud" around us.

I know that individually none of the above solutions are amazing or revolutionary. But the cumulative impact of the incremental improvements would be life-changing and perhaps even planet-saving.

To go a bit further ... and maybe even more practical (at least from the business world perspective) ... consider the many office workers in the world today.

Think about it ..... they do need to go to the office, they do not build physical things or manage plant equipment, and do not need to directly collaborate on a continous basis. They could just as easily meet in virtual environments. There would be huge savings in avoiding commuting to the office (time, gas, vehicle costs, environmental impact). Office space could be scaled back (infrastructure savings). As long as the phones are diverted you could have meetings in the virtual world. Documents can all be electronic etc.

The work force could be further dispursed without the need to regularly come into work. Distance would no longer be a factor. Also, as long as the Virtual Environment is setup like a normal office ..... managing the people should not be too different. Although new tools may be needed. I'm sure Microsoft or some other entreprenurial entity would be happy to fill any need.

It would also be possible to build web monitoring of Factories etc into the virtual world using web cameras etc. You could visit a factory in another state or country to see how its operating without actually going there.

Robots could be used as avatars in the real world to interact with the virtual world. A virtual avatar linked to a robot seeing all the same things that exist in the real environment could interact in industry's with manufacturing etc. This is particularly handy in dangerous industries, as it is currently used in bomb disposal but linked to a virtual environment.

That's just a bit of "brainstorming" as a starter. Now ... what does YOUR imagination say? Leave a comment and share!

Business VoIP Solution .... What Does It Really Mean?

A business voip solution is an alternative to a traditional switched-voice service. Instead of using dedicated trunks between PBXs, you can share the bandwidth with your data services, making better use of the available capacity. With prioritization techniques, you can ensure that your voice traffic gets through the network in the appropriate time to maintain voice quality.

For a typical internet telephony application, assume that you have a number of offices that need to communicate with each other. The telephony traffic can be carried across the internet or across a third party network or VPN .... and be delivered to the distant end without using the PSTN or a leased line. It depends on what you want to achieve.

You can have an IP PBX for internal use only, with all your staff connected over the same LAN infrastructure as they use for data - separating voice and data traffic by using VLANs, with external access to the PSTN. If you have more than one office, you can link them using leased T1 or DS3 bandwidth lines, metro ethernet or xDSL in the same way as you would for a data service - VoIP is, after all, voice packets being transported in the same way as data packets.

As the VoIP traffic is handled just like any other data traffic, it doesn't really matter where your PBX is located, just as you can have a hosted server for your data, you can have a hosted server for your VoIP. You can have the IPPBX in your office or in a data centre, it makes no difference.

SIP trunking? Now there's a buzz phrase. SIP is a signaling protocol in the same way as H.323 or QSIG are protocols. SIP is a standard and lots of manufacturers intrepret the standard in their own way but, fortunately, lots of the SIP interpretations work together.

Session Initiation Protocol works over IP and should enable replacement of PSTN trunks by the internet. So, again, voice traffic can go over the internet instead of the PSTN with a resulting saving in cost.

Don't forget though, that the internet is unregulated and unmanaged. Voice traffic is sensitive to packet loss, delay and jitter (variation in delay). So at peak times ( just after school in the evenings and on weekends) your voice traffic may break up. Meaning it would be worthwhile looking at having a managed network service to guarantee the quality of your voice traffic for those critical calls.

VoIP works between products from the same manufacturer and sometimes it works between manufacturers. Some manufacturers have produced interfaces and some third parties will provide gateways to enable collaboration. My recommendation would be to check first.

Hardware? Go for something that lots of other people use.. Cisco, Avaya, Alcatel. Cisco started in data and moved into voice. Avaya and Alcatel started in voice and moved into IP data. They have their own ways of implementing VoIP. Cisco is only VoIP. Avaya and Alcatel can give you a migration from traditional voice, reusing handsets from their analog and digital services adding IP telephones as the transition continues. IP telephone handsets are expensive.

For more help in designing a business VoIP solution for your network .... I recommend taking advantage of the services available here:

Business VoIP Solution

Their assistance is free and and they often can find special deals from providers in your specific location.

What Is Your Disaster Recovery Plan If Your Legacy Communications And Network Platforms Were To Fail?

This is a question that comes up frequently among users of legacy and secondary market end-users. I find that it is crucial that the end-users receive the necessary technical support and warranties to keep their network up and running at all times. If there ever is downtime, they should have a hardware provider that can overnight new equipment so as to minimize inefficiencies.

At some point in time your legacy systems will no longer be supported and a failure will severely impact your operation.

The lack of support could be technical, the company no longer manufactures the product and parts are harder to find or no longer available, or the lack of support could be no people with the skills necessary to maintain the equipment.

That is one reason that you need to ensure that your plan is tested on a regular basis. These support issues should become evident during the test. If you are using a third party recovery service provider, they may tell you that they no longer support the legacy system. They may make suggestions on what they can do to accommodate your systems, but you may find that the cost is too high. It may actually be better to do an upgrade.

You also need to make sure that any contract that you sign with a recovery service provider will let you upgrade at little or no cost to the contract price. You also need to include a clause that will let you know when the recovery service provider will withdraw support so that you can plan your response.

BCP is not a project, so you need to make sure that the plan is reviewed and tested regularly so that problems can be identified and corrected before an emergency or disaster event occurs.

A disaster recovery plan should include all key componets necessary to run an enterprise. From a network and communications point of view this may include planning for redundancy not only in the Local Area Network but with the carriers. I had a WAN at one client where I had automatic roll over should the main gateway fail at any point. Many enterprise have backup plans with secondary carriers and can re-point 800 numbers within minutes.

DRP planning is different for all enterprises .... but to be effective needs managment support from the top. I suggest you read the following article as a general overiew on disaster recovery.

Disaster Recovery Planning

What You Need To Know About A MPLS Network

There's no one "MPLS Service". A lot of what you get will depend how the service provider has Engineered and built their core network. Bear in mind that many carriers don't own the whole network, but will piece together a service from other carriers networks, or will interconnect with other carriers to extend their reach.

Cell-mode MPLS was mentioned: basically this is ATM which has been retro-fitted with MPLS. Be careful with this for VoIP applications because it can use bandwidth very inefficiently.

MPLS can support QoS, but many services aren't engineered with this, or only with very basic prioritisation. Also the services are very often structured to reduce the potential complexity and to ensure the network can cope. Bear in mind a typical MPLS router can only carry a percentage of "high-priority, realtime" traffic. If everyone sends all their data as high priority then the benefit is lost, and the network may suffer. Usually QoS is provided as a small number of service classes, typically 3 or 4.

The biggest bottleneck in any such service is normally the tail circuit to each of your premises. If you move from a T1 mesh to a MPLS service then you will likely find that some sites need more bandwidth than others. Tracking the requirement for this bandwidth is usually your problem, although the service provides may give you some reporting tools to assist with this. I would avoid service providers who cannot offer this as it will make it very difficult for you to manage your bandwidths.

If you factor in multiple service classes then your management of these tail circuits gets more complex as you no only have to work out how much bandwidth is required for each tail circuit, but how much of it should be reserved for each service class.

Regarding resilience, within the service providers core, the service is normally highly resilient to failures. However, when failures do occur, very often (depending on how the service is engineered) the rerouting can take a second or two. During this network reconvergance you will lose packets. Depending on the protocol your traffic uses this can be unimportant or devastating. For instance, some VPN and VoIP services don't survive this well.

Normally resilience is not automatically provided all the way to the customer. Typically you will have one tail circuit and one router at each site. If either fails (or if the Service provider's PE router has problems) you will lose service to the site, totally.

If this is an issue, you need to factor in dual connections. There's multiple ways of doing this, and different service providers will offer different options. Make sure you get your Network Engineer involved as the devil is in the detail here, and some options which sound like they provide a fantastic level of resilience may not be as good as they sound, depending on how your internal network is configured.

And, of course, the key to all of this is SLAs: what do they offer? What happens if they break them? How do they report them to you?

Generally speaking, MPLS services are a great way to run a multi-site data network including VoIP services. I have seen many carriers and their customers doing this successfully for years.

Strictly speaking MPLS does not provide QoS. QoS is done by prioritising traffic, and most IP routers, even those on the backbone of the Internet, can do this. The difference is whether they are configured to do this or not. In an MPLS network MPLS is provided by standard router features. MPLS technology (specifically Traffic Engineering) gives the carrier better control over how this traffic is prioritised and routed (and restored in case of network fault). All this does is give them the confidence to support SLAs.

As I mentioned, "QoS" is provided as a set of "service classes". Typically these are things like "real-time", "high-priority" and "everything else". Mapping actual traffic into these classes can be done in a few different ways, but this is largely up to you to control. For instance, you could quite easily put web-browsing traffic into "real time" although this would normally be a dumb thing to do.

I would suggest the case for MPLS in terms of performance, cost and continuity against 'traditional' or 'legacy' data networks is now pretty robust, i.e. MPLS provides significant advantages in all 3 areas.

The key considerations when migrating include provider selection, access media (e.g. using Ethernet rather than SDH/SONET), the decision on procuring a managed or unmanaged service (often called wires-only) and the providers ability to map their CoS/QoS to the applications you need to support. This is especially important if you are operating any proprietary applications.

There is also an increasing trend to use WAN Optimisation/application management solutions either as a value added service from the provider or from an alternative integrator or indeed doing it in house. This is important say for voice or applications such as CITRIX.

MPLS providers also now offer a whole suite of value added services such as integrated internet, managed network based firewalls and remote user support. If these are important to you make sure the providers demonstrate how this is achieved.

In selecting your provider ensure they have good geographic coverage in your areas and experience within you market segment. I always recommend taking up 3 references. Equally I think it is wise to understand how important a client you will be to the provider; it's all well and good using the market leader (say according to Gartner) but you'll often get a better service from a provider who values and really wants your business.

For FREE assistance designing the right MPLS configuration for your network .... AND sourcing the most cost effective provider ..... take advantage of the help available via:

MPLS Networks

Moving To MPLS Network Architecture From WAN Architecture

You're tasked with investigating if moving your company's network from a point-to-point T1 WAN architecture to a MPLS architecture makes business sense.

The easy answer here is yes.

But .... why?

In looking at changing your architecture from Point to Point to an MPLS type of network I suggest starting with the business requirements and tying your network requirements to the business needs. In this manner, you'll have clear business outcomes which you can negotiate back with you're business when you start getting the cost vs benefits discussion in developing your business case for investment.

A robust Total Cost of Ownership model will be needed to understand what the TCO will be gonig forwards. Also, I'd suggest developing a strong understanding of the costs of doing nothing and also the potential savings or new revenue opportunities for your business so you can develop a Net Present Value (NPV) of your network options.

I would also recommend looking at the interfaces you're looking to support in the network. MPLS does enable you to have a common protocol across all your networks and you can effectively establish an MPLS Cross Connect in your network. This will rely on how your local service providers will provide MPLS services to you, if at all. So you would most likely need to purchase either point to point or point to multipoint based transmission services from your provider. We're seeing that many enterprises and service providers are heading towards ethernet ubiquity as a service interface and then offering multiple services ontop of the ethernet interface.

VoIP works well across an MPLS type of network, however it does depend on the services that you purchase off your service provider. As you're looking at MPLS, then I assume that you're looking at buying straight transmission services and then you'll use MPLS to aggregrate traffic into your WAN links. Hence, you're business case is going to be driven by arbitrage opportunities so capture as much traffic as possible onto your network and apply QoS at the edge.

From a QoS perspective, ensure that you can also apply policing to the traffic that goes onto your WAN traffic. I recommend applying Heirarchical QoS as this will enable you to dynamically share the bandwidth in your WAN links.

As you can tell, there are lots of issues and questions that need to be addressed so I'd suggest working closely with some trusted partners and driving towards an outcome based business drivers and commercial outcomes.

We have worked with many customers that have migrated to MPLS from old school point to point. There are a few reasons our customers did this, but let me assure you the #1 reason was cost. A good competitive carrier will offer an MPLS solution that is sometimes less costly than the old point to point type solution ..... with most of the same or more functionality.

But there are some factors:

1. If the point to points are crossing state or lata boundaries ..... or are fairly separated by miles .... you should enjoy considerable savings.

2. If you get a carrier that bundles MPLS with an Integrated Access type solution you will save big money (combination of voice, internet & MPLS delivered on one T1 with quality of service).

Here's some additional points that might help you.

1. Is it redundant? Yes, depending on how you design your network (we can help you of course) sites can network with each other over your wide area network for disaster recovery/ redundancy. Unlike the traditional point to point architecture where you might only be as strong as that single link. We helped a huge national company with a migration from point to point and frame relay to MPLS. The big reason was with so many sites there was an outage almost every day. The network was designed with redundancy as the main driver.

2. Does it work as well? It depends who you ask. Are you talking to a salesman? He will say yes. Let me give you my "consultant" opinion. It works ALMOST as well but there are so many benefits to MPLS that typically motivate a customer to change. For example, if you have Quality of Service (QoS) sensitive applications running accross your WAN then you should consider MPLS. MPLS is a private networking technology similar to the concept of Frame Relay in that it is delivered in the "cloud". The primary difference with MPLS is that you can purchase quality of service for applications across your WAN. During the provisioning process the carrier (or your agent-wink wink) will interview you in order to determine which appications are important to your business, they will then build a QoS template to service these applications on your WAN. These applications will be given priority over all other traffic in times of peak load. MPLS is by far the most costly solution between Frame Relay, VPN and MPLS .... but is the only technology that will support QoS!

But let's put the salesman aside and remember one thing. With MPLS we are using the carrier's private network which is infinitely better than creating your own VPN. But because of some "overhead" and the belief that all carriers over-subscribe somewhat I am convinced that it's ALMOST as good.

So if it's almost as good would it be worth migrating if you could enjoy cost savings & redundancy??? Maybe. But these are the things that are making MPLS the hot ticket now.

3. Does this work well with VOIP? Sure. You can get QoS like I stated above.

MPLS is an IP-based framing technology (at OSI layer 3) that inherently meshes your WAN (this is the redundancy you refer to above). MPLS has a feature called QoS or quality of service. This feature allows both your CPE router and the carrier's network to prioritize data based upon your settings or preferences (carrier's level of support of QoS can be broad) and gives you more "bang for your buck" with the bandwidth that you select for the local loop going to each office. MPLS is made for VoIP like RC Cola is made for Moonpies. Because true "toll quality" VoIP requires prioritization across a carrier network, you (or your provider) can tag VoIP traffic with high priority to easily address the jitter and latency sensitivity inherent in the service.

The other huge advantage you have is that you can add locations with a simple routing table update and maintain a fully-meshed architecture, where with Point to Point circuits, you would have had to add a separate circuit to each location you want to interconnect, making MPLS more and more cost friendly the more locations you add.

Finally, MPLS allows you (or your carrier) to configure network objects (such as servers, VPN concentrators, and Network-Based Firewalls) as nodes on your MPLS network. For instance, with a properly deployed Network Based Firewall, you can provide all of your locations with an internet connection over your MPLS network that doesn't rely on a single location to aggregate the traffic. Some carriers even offer redundant firewalls, meaning that you have redundant Internet connections fully meshed giving you more potential up-time in the case of a single failure on your network.

What do you need to consider? In my experience, the biggest things to keep in mind are:

- Stay away from MPLS enabled Frame/ATM networks with committed access rates (CAR), this committed access rate is often a lower bandwidth than your local loop bandwidth, which can degrade your quality and quantity of bandwidth across a carrier's network (its always in the fine print).

- QoS at the "Edge" and across the "Core" - choose a carrier with both.

- QoS recognition across the carrier network - some carriers will allow customers to mark packets with priority, but will not recognize and uphold that priority, don't fall into this trap.

- SLA guarantees - be sure to choose a carrier that provides acceptable service level agreements for the type of service you plan to push across your network

- Customer Service and dedicated sales rep - you want someone you can reach out to with questions that you can trust - this is the most difficult thing to find.

For FREE assistance designing the right MPLS configuration for your network .... AND sourcing the most cost effective provider ..... take advantage of the help available via:

MPLS Networks