Individuals, businesses and organizations connect to the Internet in a variety of ways depending on a multitude of factors. Connecting to the Internet can be through one or a combination of the following: subscribing to the Internet through an Internet service provider, procuring IP transit, and peering. On this page the primary focus is on IP transit, especially for those interested in procuring the service for their organizations.
IP transit is when a large network (or upstream) provider allows Internet traffic from an organization, business or smaller Internet service provider to pass or “transit” through its network. When your organization procures an IP transit port, it will not only have access to the Internet but it can also allow others to access the Internet through the organization’s network. Furthermore, your organization will have access to the provider’s network and the networks of provider’s peers and other IP transit clients.
Unlike dedicated Internet access where anyone can sign up with an Internet service provider, an organization requires an Autonomous System Number (ASN) to procure IP transit service. An ASN identifies a network on the Internet and defines a group of routing prefixes that maintains a unique routing policy for the network. Every network on the Internet has an ASN and an associated block of IP addresses assigned by the Internet Assigned Numbers Authority (IANA). Any organization wishing to offer Internet access to its employees, members or customers may be granted Autonomous System status and a block of IP addresses after providing a justification of how the IP addresses will be used. Because there is a finite number of IP addresses (IPv4 and IPv6) available, the authority ensures that they are used efficiently. ASNs can be requested from regional authorities (given below) or through the upstream provider.
The following organizations typically buy IP transit:
Organizations wishing to apply for autonomous system status and IP address blocks for their networks should contact the relevant authority in their region:
Blocks of IP addresses are allocated by the regional authorities given above. An IP address block is divided into two parts, a part identifying the network on the Internet and a prefix that defines the size of the block. If an IPv4 block is presented as “192.168.2.0/26”, for example, “192.168.2.0” identifies the network and “26” is the number of bits in the network. Since IPv4 is 32 bit, only 4 bits (32-26 bits) are available for IP addresses. Thus, this is a small block with a total of 16 IP addresses. If the block was “192.168.2.0/2”, then there will be 30 bits available for IP addresses translating to 1,073,741,824 IP addresses – a very large block indeed. The smaller the number after “/”, the larger the block and vice versa. The block size alone is simply given as “/n”, where n is the number of bits in the network. The charts below, the so called Classless Inter-Domain Routing (CIDR) Charts, give more details for both IPv4 and IPv6 blocks.
There are two cost components in IP transit procurement – the provider service charge and the ASN and IP addresses charge. The provider’s charge will include a one-time installation charge and a Monthly Recurring Charge (MRC). The MRC may be fixed for an agreed upon data rate, also known as the committed data rate, and is payable even if the actual bandwidth usage is lower. However, if usage goes over the committed data rate, additional charges will be due. Alternatively, the MRC can also be made to vary with bandwidth usage – a model most suited for organizations with low bandwidth requirements. Most providers price their service per unit of bandwidth such as cost per Megabit/s. The higher the bandwidth subscribed to and the longer the contract term, the lower the cost per unit. For example, a company that subscribes to a committed data rate of 10Gigabit/s with a contact term of 5 years will be charged significantly less per Megabit/s than one that subscribes to 10Mbit/s with a contract term of 1 year.
ASNs and IP addresses can be obtained directly from regional registries for Internet numbers – American Registry of Internet Numbers (ARIN) in the case of North America. Some IP transit providers may be able to offer IP addresses to their clients. Contractual terms may vary depending on how your organization is classified – end user for organizations that need IP addresses for internal use and Internet Service Provider (ISP) for organizations that need IP addresses for distribution to external clients. ASNs and IP addresses also have two cost components – an upfront cost depending on the IP address block size and an annual maintenance cost. In North America upfront costs range from $250 for /24 IPv4 (/40 IPv6) block or smaller to $256,000 for /6 IPv4 (/4 IPv6) or larger. In addition to the fixed charges, maintenance fees of $150 for any ASN, $150 for any IPv4 address block, and $150 for any IPv6 address block are payable annually.
If you request a quotation from multiple providers based only on the bandwidth, it is more than likely that you will receive you will receive a wide range of prices. The temptation for many small organizations is to jump on the least cost solution without paying attention to the features responsible for the cost disparity. It is strongly recommended that you review some of the important IP transit features to make an apple to apple comparison of pricing from different providers. Moreover, you have to carefully match your business requirements with what each provider is offering. Following are the most common features of IP transit services.
Most, if not all, IP transit providers require that you commit to a contract term, typically 1 to 3 years. Some may even offer terms of up to 5 years. You have to carefully weigh the pros and cons for your organization before settling for a contract term. You may contact FiberGuide to help review your service and offer guidance on the most suitable term. The longer the term, the lower the MRC and your monthly charge may be locked in to a fixed price, protecting you against any potential future price increases. On the other hand, if the service does not turn out to be what you expected, it will cost you to get out of the contract. It is advisable to start with a one year contract term and then extend to a longer term if you are happy with the service.
New organizations, especially ISPs, prefer to start small and increase their bandwidth as their own clientele grows. Scalability is the ability of the network bandwidth to be increased by small amounts without significant changes to the infrastructure. Ethernet, for example, is more scalable than SONET or SDH.
Most IP transit providers offer a wide range of interfaces to connect your network to theirs. The onus is on you to opt for the most suitable interface that meets your organization’s requirements. Unless there is a compelling reason for opting for a TDM based interface, such as existing TDM network, conventional wisdom is to opt for an Ethernet based interface. In addition to its ubiquity, Ethernet is flexible and is easily scalable. You can start with lower bandwidth such a 20 Megabit/s and be able to upgrade to 10 Gigabit/s without significant changes in infrastructure. Moreover, the industry is trending towards Ethernet and TDM may soon be obsolete. The following interfaces are readily available from most providers:
The ever expansion of the Internet has resulted in IPv4 addresses running out. IPv4 is 32 bit with a total of about 4 billion addresses which can no longer meet the demand. IPv6 was introduced to provide sufficient addresses so that we can never run out of IP addresses again – at least not in the foreseeable future. IPv6 is 128 bit which translates to hundreds of trillions of possible IP addresses. However, until the whole Internet completely shifts to using IPv6, IPv4 and IPv6 have to co-exist. A dual stack network is one in which all the nodes are both IPv4 and IPv6 compatible.
Service level agreement the commitment that the IP transit provider makes to their client about the service. It includes that compensation promised by the provider in the event that certain aspect of the agreement are not adhered to. For example, if there is a downtime greater than that specified in the SLA, the provider may waive part of the MRC. Quality of service is one of the most important component of a SLA and it may include, among other attributes, availability, latency, jitter and packet loss.
Availability is the probability that the network will be up and running at any given time. Availability is specified as a percentage of the time that the network is up and running over a period of time. A 99.5% availability translates to a total network downtime of about 2 days in a year. A 99.9999999% (9 9s) availability translates only to about 32 milliseconds of downtime in a year. However, for IP transit providers typically offer 99.9% to 99.999% availability and only a few can achieve the upper limit.
Latency (typically in ms) defines the time taken by a data packet to travel from the sender to the receiver. It depends on the distance travelled, the transmission medium and the equipment (such as routers) through which the packets travel. If the medium is optical fiber then there is a 4.90microsecond/km (0.0094milliseconds/km) contribution to latency which can add up significantly if packets have to be routed through too many hops. If geosynchronous satellite systems are part of the IP transit network, then there is at least 540 milliseconds contribution from the data packet round-trip to and from the satellite. Certain applications such as videoconferencing, VoIP, gaming and high frequency trading are very prone to latency. In VoIP, for example, latency manifests itself as a delay between the time you talk in a handset and the time you hear a response from the called part. In a VoIP conversation, latency starts to be noticeable at about 250ms and it starts to get irritating at 300ms. At 500ms it becomes a total nuisance.
Jitter occurs when latency changes with time. With a fixed latency, two parties to a conversation can come to terms with the delay between them and eventually have a smooth conversation. With jitter, however, the latency changes randomly over time, making it difficult to communicate.
The objective of any network provider is to deliver 100% of data packets originating from a sender to a receiver. However, inevitable network errors will cause some packets to be lost. Packet loss is defined as the percentage of the number of packets lost to the number of packets transmitted. Network technics such as forward error correction are used to compensate for the lost packets but this can come at a penalty in time – in other words packet loss slows the network. When packet loss is too high, it can manifest itself as a lost connection.
The Internet is usually described as a giant network of networks connected to each other using the TCP/IP communication protocol. At the core of the Internet is a very high capacity optical backbone connecting continents and regions with submarine cable and terrestrial long haul cable respectively. The optical backbone is a network of the world’s largest networks (Tier 1 networks) seamlessly connected together into one network. The Tier 1 network providers peer with each other and allow IP traffic to flow from one network to another free of charge. A Tier 1 Internet provider can therefore be defined as one that connects to the Internet through unpaid peering. The peering arrangements between Tier 1 providers is also referred to as settlement free inter-connection. There are only about a dozen Tier 1 network providers, although there are many more that may claim to be Tier 1. If they don’t access the Internet by peering ONLY then they are not a Tier 1 provider in the strict definition of the term. Level3 (now CenturyLink), AT&T, Verizon, Cogent, Global Telecom and Technology (GTT), Telia Carrier, NTT Communications, Tata Communications and Telecom Italia Sparkle are some of the Tier 1 providers.
The following is true of all Tier 1 Internet providers:
Headquartered in the USA, AT&T (American Telephone and Telegraph Company) boasts of one of the largest global backbone fiber optic network. The network transport about 5 petabytes (that is 5 million Gigabytes!) of data on any given business day. The backbone network consists of over 500, 000 route miles.
CenturyLink (includes Level 3)
CenturyLink is an American company and is the second largest provider of networking solutions to businesses. Centurylink’s global Tier 1 IP network is over 500,000 route miles. The backbone has a capacity of 20.9 Tbps and they can offer 100 Gbps bandwidth per optical wavelength. CenturyLink has more than 40,000 on-net buildings (and growing) from which clients can easily access IP transit and other connectivity solutions.
Based in Japan, NTT (Nippon Telegraph and Telephone) is one of the 5 largest global IP network providers. NTT has 2.5 Terabits of private peering with other Tier 1 network providers and has points of presence (PoP) in 190 countries.
Tata communications (formerly VSNL) is an Indian based Tier1 network provider. Its backbone network consists of 500,000 km of submarine optical fiber, 210, 000km of terrestrial optical fiber, and 600 PoPs in more than 200 countries.
Telia Carrier (formerly TeliaSonera) is a leading global IP network provider based in Sweden. In May 2019, Dyn ranked Telia Carrier the number 1 Internet backbone in the world. Telia Carrier has more than 265 PoPs serving IP customers in more than 110 countries. Their IP backbone network is 65,000 fiber km covering North America, Europe, Asia and the Middle East.
Verizon Enterprise Solutions
Verizon, an American company, operates an expansive network to deliver IP services all over the world. The 500, 000 route km network traverses 150 countries in six continents. Verizon is one of a few network providers that owns both IP and analog voice infrastructure in many locations around the world.
Zayo Group owns and operates a 130,000 route mile network (translates to 12.5 Million miles of optical fiber) in North America and Europe. The network includes connectivity to thousands of data centers, carrier exchange points, financial exchanges and other enterprise locations.
The following is true of all tier 1 Internet providers:
Tier 2 providers peer with other Tier 2 providers and, in addition, buy IP transit from Tier providers. Tier 2 providers include large Internet service providers such as incumbent and competitive local carriers. They include Comcast, Google Fiber, Liquid Telecom Zimbabwe and British Telecom.
Tier 3 providers connect to the Internet exclusively by buying transit from Tier 2 or Tier 1 providers. They consist of small Internet service providers, enterprises and institutions.
Whether you are an enterprise with large offices, a financial services company, a technology company, an ISP, a cloud service carrier or a National Research and Educational Network (NREN), you can benefit from our IP transit service. Powered by the world’s leading IP transit carriers, our service helps you connect your network to a Tier 1 IP network provider. By accessing a Tier 1 carrier network, you seamlessly become part of the global Internet with access to the full Internet routing table. A partial list of cities with PoPs is provided at the bottom of the page.
As top tier network providers, our carrier partners offer a lot of value to clients around the world. In addition to each individual provider having an expansive global network, their networks are peered together to give a seamless global network.
Top tier network providers’ data run on all optical fiber networks with a limited number of hops. This guarantees low latency IP transit for applications such as voice, banking and other latency sensitive applications. All providers offer services with guaranteed quality of service (QoS).
Each of our providers offer network redundancy or alternative paths on their networks. This means that in the event of a catastrophic failure caused by a cable break or other events, your IP transit traffic is seamlessly routed through an alternative path.
Our providers are connected to major data centers around the world. In addition to offering easy access to their networks, they also have easy access to all major content distribution networks (CDNs), over the top (OTT) providers and the private cloud. Whether you or your clients want to access Amazon Cloud Services, Microsoft Azure, Google Cloud Platform or any other cloud services, our providers easily let you do just that.
Our providers run on both IPv4 and IPv6. If you are an ISP, you want to ensure that your IP transit network provider has IPv6. IPv4 will soon be obsolete and you want to ensure that you will not run out of IP addresses when that happens. When shopping for an IP transit provider, IPv6 is one thing you must look out for.
We are a one stop source for IP transit solutions, and we partner with most major providers. You can leave the daunting task of checking with multiple providers for a suitable solution. We have easy access to them and can get you the information or RFQ responses in a breeze.
Our carrier partners have millions of combined route kilometers of optical fiber around the globe. This means a wide global coverage options for you to choose from. Whether you want to connect locations in Honolulu, Mombasa, New York, Canterbury, Christ Church or thousands of other locations around the globe, we have you covered. Our partners have tens of thousands of combined points of presence (PoPs).
While each of our providers have bandwidth flexibility owing to the beauty of carrier Ethernet they use, we offer even greater flexibility by selecting from all their services for our clients. We endeavor for find you port speeds ranging from 10 Mbps to as high as 100 000 Mbps for your IP transit solution. FiberGuide also offer quotations for wavelength services and business Internet providers.