The modern computer is a device with many capabilities. It is quick, portable, and, most importantly, it can communicate with other computers. An Internet protocol address (IP address) is the numerical reference assigned to each computer on a network. A network can range from a small private network to the greater public network of networks known as the Internet. An IP address is necessary to allow those billions of devices to communicate with one another, so messages can be sent and responses sent back. The devices sending the messages and all the intermediaries routing the messages all over the world need to know where to send those messages. Networking is complex. It is the result of a simple concept being built upon by experts for over 40 years and modern computing relies on its success. The speed of communication between computers on private networks and over the Internet has become as essential as the speed of the computer itself.
IPv4 (Internet Protocol version 4) is the IP version implemented in 1981 and is still in use today (Johnson, 2011). IPv4 has served its users well over the years and is still the more popular of the two protocols used now. It is not without its faults, however. The most critical issue with IPv4 is that it was implemented before anyone had any idea how prolific computers would be just 30 years later. IPv4 uses 32-bit addressing, which leaves approximately 4.3 billion IPv4 addresses available for use (West, Dean, & Andrews, 2016). As of 2018, the number of Internet connected devices worldwide was 17 billion and rising (Lueth, 2018). There are now more than triple the number of devices than there are IPv4 addresses to connect them. IPv4 has stayed relevant with modern technology with the assistance of NAT (Network Address Translation).
NAT allows a private network of IP addresses to exist behind a single, public IP address. The private address is stripped from each packet by the router or firewall before being sent out on the Internet and the address of the NAT enabled router or firewall is included in the packet instead. This prevents some amount of unwanted traffic because outsiders never know the actual address of a device. Individuals or organizations are able to lease a small number of addresses to network the multitude of devices they may need to connect. NAT is not the only way that organizations are stretching the use of IPv4 addresses.
CIDR (Classless Inter-Domain Routing), also called supernetting, is an additional tool used with IPv4 addressing. Currently, many large organizations have leased class A or B addresses which other individuals or organizations then cannot use. Some of those leases were acquired when IPv4 first came about, before running out of addresses was a concern. CIDR is used to prevent organizations from amassing swaths of unused IPv4 addresses and to prolong the use of IPv4. In classful addressing, IPv4 addresses are separated into distinct classes based on the number of bits in the subnet mask. The network ID of a class A address is 8 bits, which means there are 16 million addresses in each network. Class B addresses use 16 bits and have 65,000 possible IP addresses. Class C uses 24 bits and allow 254 possible IP addresses per network. As the number of possible IP addresses in a network decrease, the number of possible networks increases (West, Dean, & Andrews, 2016). This is where classless addressing is beneficial. CIDR allows the network ID to be a variable length, such as 26 bits, which enables the allocated number of IP addresses to more closely fit the actual size of the network. The network ID can be used to split the network into smaller network segments. A router is given an IP address that matches the network ID of the nodes in its subnet. This way, with higher-level routing, routing tables only need the IP addresses of other routers in order for traffic to be directed to the correct network segment.
The benefit of IPv4 depletion is that it will force implementation of a newer and more versatile protocol, IPv6. As servers and other computing devices reach their natural end-of-life, they will likely be replaced by IPv6 compatible devices. Reasons companies have for not switching include money, time, and training. At a certain point, the benefits of IPv6 will outweigh the cost of switching.
IPv6 addresses are 128 bits long. There are about 3.4 x 1038 possible IPv6 addresses (West, Dean, & Andrews, 2016). It is not just the massive number of IP addresses that makes IPv6 special. It also has a higher efficiency when routing packets. Another benefit is the ability to connect a device to multiple networks without requiring multiple network interface cards on the device. IPv6 also replaces broadcast communication with multi-cast. Broadcast packets are indiscriminately sent to everyone on a network. With IPv6, hosts that are not interested no longer have to process the data in a broadcast packet. Multi-cast packets are specifically assigned to each destination host, meaning packets are sent only where they are intended, which saves on throughput (Dargin, 2017).
IPv4 was designed before network security was a serious consideration. IPv6 provides secure options, such as support for IPsec (Dargin, 2017). IPsec is “a suite of protocols that provide data integrity, confidentiality, and authentication” (Caicedo, Joshi, & Tuladhar, 2009). With IPv6, the Dynamic Host Configuration Protocol (DHCP) server is no longer required for a node to be assigned an IP address. The node can now use local information to determine its own address. This again prevents waste on a network if devices do not have to periodically request a new IP address each time a lease expires. Another benefit of IPv6 is that it allows ISPs (Internet Service Providers) to reduce the size of routing tables by encapsulating IP addresses for customers under a single prefix. The ISP does not need to list every IP address in order to route traffic and is able to save time by routing between fewer addresses (Dargin, 2017).
IPv6 addresses on a network can be protected in a way similar to how NAT protects IPv4 addresses. Mike Meyers, author of CompTIA Network+ Certification All-in-One Exam Guide, states, “a proxy server sits in between clients and external servers, essentially pocketing the requests from the clients for server resources and making those requests itself” (Meyers, 2015). The proxy server makes requests on behalf of its clients, which means external devices attempting unsolicited traffic will only see the proxy server and will therefore send to the proxy server rather than the intended target. If the proxy server does not see an existing connection between the solicitor and a computer on its network then the uninitiated traffic will be dropped. Proxy servers allow clients to hide their identities from outside devices as well as provide numerous other services. A proxy server operates at a higher level with greater capabilities than a NAT device.
One might wonder how IPv4 and IPv6 can coexist and still allow devices on networks using different protocols to communicate. The answer is tunnels. Networks using both IPv4 and IPv6 are generally configured to use both protocols without requiring a tunnel. However, many networks still use only IPv4 and have no such configuration. Most areas of the Internet are not configured to use both and require tunneling for IPv6 traffic. 6to4 is a tunneling protocol that allows IPv6 traffic to travel through an IPv4 network. The IPv6 packet is encapsulated inside an IPv4 packet. 6in4 is one of two protocols that allow IPv6 traffic to travel through a NAT. Teredo is the other option. One way for IPv4 traffic to travel over a network configured to use only IPv6 is called 4to6. 4to6 is a process that involves encapsulating IPv4 traffic within IPv6 (West, Dean, & Andrews, 2016).
Tunneling is beneficial in that it allows IPv4 to continue as the primary protocol while IPv6 is gradually implemented. The disadvantage is that forcing IPv6 traffic to act like IPv4 traffic over the Internet stunts the capabilities of IPv6. The difference in efficiency of IPv6 versus IPv4 networks may seem insignificant when implemented on a network of smaller scale. Full implementation over the Internet would see drastic improvements in the speed of transmission of data. Network security would also improve because filtering devices would be able to read and sort more data if that data is not hidden behind tunneling protocols.
IPv6 implementation is increasing rapidly in industries where speed is everything. Internet Society touts, “over 25% of all Internet-connected networks advertise IPv6 connectivity” (Internet Society, 2018). Mobile carriers are leading the pack with implementation rates upwards of 90 percent. Reliance JIO of India is at 87%. Verizon Wireless is at 84%, Sprint at 70%, T-Mobile at 93%, and AT&T Wireless is at 57% (Internet Society, 2018). Some companies are even looking to shut off IPv4 in the near future in order to simplify their networks. Facebook is already in the process of turning off IPv4 within their data centers. Microsoft and LinkedIn have stated their intention of doing so as well. Broadband ISPs are also in a group of high implementation. Comcast is at over 66% deployment, Sky Broadcasting is over 86%, and AT&T is almost 66%. Reliance Jio is at 86% and has over 237 million IPv6 users. The main group dragging their heels with implementation are enterprise networks. Those private companies that do not earn primarily based on how efficiently they can process data are less motivated to make a change with such high overhead cost. The cost of continuing on will eventually be enough to warrant a change (Internet Society, 2018).
The price of IPv4 addresses is already becoming more significant. Microsoft reportedly purchased “666,000 addresses at $11.25 per address in 2011” (Internet Society, 2018). That is almost 7.5 million dollars just for IP addresses. As supply dwindles, cost is only going to increase. Before long IPv6 will be the primary implementation. IPv4 address prices may eventually decrease as demand decreases but the inconvenience of managing and setting up networks to handle both IPv6 and IPv4 will be considerable and likely more than most companies will want to deal with. There is a great deal more to be done before IPv6 is even a majority but someday it will dominate. Someday everyone will be able to appreciate the speed and efficiency of an IPv6 Internet.
References
Caicedo, C. E., Joshi, N. B., & Tuladhar, S. R. (2009, February 10). IPv6 Security Challenges. Retrieved January 27, 2019, from IEEE Xplore Digital Library: https://ieeexplore.ieee.org/abstract/document/4781968
Dargin, M. (2017, June 26). Time to Consider a Move to IPv6. Retrieved January 27, 2019, from Network World: https://www.networkworld.com/article/3203708/lan-wan/time-to-consider-a-move-to-ipv6.html
Internet Society. (2018, June 6). State of IPv6 Deployment 2018. Retrieved February 2, 2019, from Internet Society: https://www.internetsociety.org/resources/2018/state-of-ipv6-deployment-2018/
Johnson, B. (2011, February 4). The Internet Just Ran Out of Numbers. Retrieved January 21, 2019, from MIT Technology Review: https://www.technologyreview.com/s/422612/the-internet-just-ran-out-of-numbers/
Lueth, K. L. (2018, August 8). State of the IoT 2018: Number of IoT devices now at 7B – Market accelerating. Retrieved January 21, 2019, from IoT Analytics: https://iot-analytics.com/state-of-the-iot-update-q1-q2-2018-number-of-iot-devices-now-7b/
Meyers, M. (2015). CompTIA Network+ Certification All-in-One Exam Guide. New York: McGraw-Hill Education.
Shaw, K. (2018, September 27). What is IPv6, and why aren’t we there yet? Retrieved December 16, 2018, from Network World: https://www.networkworld.com/article/3254575/lan-wan/what-is-ipv6-and-why-aren-t-we-there-yet.html
Tsirtsis, G., & Srisuresh, P. (2000, February). Network Address Translation - Protocol Translation (NAT-PT). Retrieved January 21, 2019, from The Internet Society: http://www.rfc-editor.org/rfc/pdfrfc/rfc2766.txt.pdf
West, J., Dean, T., & Andrews, J. (2016). Network+ Guide to Networks (7th Edition ed.). Boston, Massachusetts: Cengage Learning.