TCP/IP Suite and its Layers
The internet is perhaps the best invention of the 20th century owing to the revolutionizing effects on the global scene for different purposes including social and corporate usage. The internet is characterized by various technologies such as protocols responsible for laying out the framework for data exchange between the respective endpoints in a given communication channel [1]. TCP/ IP is one of the communication protocols in the internet providing rules and procedures for the communication. TCP/ IP suite operates as an abstraction layer between the application programs and the respective routing and switching fabric. The growth of the internet has also culminated into new challenges straining the TCP/ IP suite in the internet suite from different fronts such as capacity. Originally, TCP protocol, was made of four layers which were later improved to add more layers making up the suite, the original TCP layers include; the application layer, transport layer, internet layer and the network access layer [2]. As a suite, TCP/ IP is a stack of layers including, the physical layer, data link layer, internet layer and the transport layer [3]. Each layer is tasked with specific functionalities to drive the working of the TCP/ IP suite such that;
The physical layer is tasked with the responsibility of specifying the attributes of the respective hardware to be used in the network such as IEEE 802.3 standards for hardware components and the RS-232 for pin connectors.
The data link layer is set to categorize the network protocol type of the respective data packets, in TCP/ IP suite, this layer is also tasked with the responsibility for performing for error control and framing, the common data link layer protocols include; Point-to-Point Protocol and the Ethernet IEEE 802.2.
The internet layer of the TCP/ IP suite is also termed as the network layer which accepts and delivers packets in the network, it is the actual layer hosting the IP aspect of the TCP/ IP suite as well as the crucial Address Resolution Protocol (ARP) and the Internet control Message Protocol (ICMP).
The transport layer ensures that the network packets get to the intended sequence error-free by swapping ACK packets of data reception and re-transmitting the lost packets. This is what is referred to as an end-end communication as is the nature of TCP/ IP connection. The TCP and the UDP protocols are the main protocols in this layer. In this scope, TCP facilitates the communication between applications by sending data appearing to have been transmitted in the form of character-by-character rather than a discrete form. In this configuration, the starting point initiates the connection to the other end. In the TCP packets, a header is attached containing different parameters including the source and destination IP to help in the routing process. An end-end connection confirms that a packet has reached the other end hence a complete communication.
The application layer is set to define the standard for internet services and the respective network applications.
Generally, this research is tailored to demystify the TCP/ IP suite and how it operates in the contemporary network system, with the changing cyberspace and the increasing demand for efficiency, TCP/ IP is facing a strain on its capabilities hence the specific needs for this research paper. The paper is thus precisely tailored to identify capacity bottlenecks, their possible causes and propose a new TCP/ IP suite model addressing these challenges.
TCP and its Role in the Suite
Specific research objectives;
- To identify the sources of TCP/ IP capacity bottlenecks
- To propose a TCP model in reaction to the causes of TCP/ IP capacity bottlenecks
Designed and developed by the Internet Engineering Task Force (IETF), TCP (Transport Control Protocol) was designed as a communication standard that facilitates message exchange between application programs and the network devices. This is a fundamental standard currently used on the internet to define the rules of communication on the internet suite for end-end data delivery. TCP sorts data in that it can be relayed between the server and the client, the end-end connection. TCP sorts data so that it can be relayed between a server and a client. It ensures that the integrity of the data sent over a network [4]. Before it relays data, TCP creates a linkage between a source and its destination, which it makes sure that it retains its live status until a communique’ begins. It then breaks massive volumes of data into smaller packets, at the same time making sure that data integrity is in place throughout the process. TCP was originally referred to as DoD model owing to the development of the networking model founded by DARPA; U.S defense agency. TCP offers ab end-to-end connection with specifications on how data must be packetized, addressed, transmitted, routed and how it will be received at the destination endpoint. These functionalities are organized into abstraction layers used to sort the related protocols as per the scope of the network; from the lowest to the highest. These layers include; link layer, internet layer, transport layer and the application layer. The link layer contains the communication techniques for the single segment of the network, the internet layer is responsible for establishing connections between the hosts while the application layer provides a process-process application data exchange. TCP/IP is a 2-layered suite; TCP as the higher layer which manages the assembling of communication messages over the internet, the lower layer which is the IP handless the addressing part to route the messages to the right destination, together, TCP and IP form what is commonly referred to as TCP/ IP suite since they two work together to handle messages and ensure they reach the intended destinations [5]. The IP (Internet Protocol) has gateways to facilitate the addressing process in which gateways check for the right addresses to forward the traffic to the intended destination despite the fact that some message packets are routed in different techniques. TCP/IP is involved in the reassembling of the packets at the end of the routing in such cases. TCP-IP is a client-server communique technique in that a connection has to be established between the two end-points before a packet is sent, the point-point communique’ allows an end-end communication making it safer in the sense that at the time there were less advancements in security threats. TCP/IP applications at the higher level are collectively said to be stateless meaning that every client request is viewed as a new request and unrelated to the previous request [6]. Being in the stateless nature, means that the network paths are free for use continuously, Other high-level application protocols include World-Wide Web Hypertext Transfer Protocol (HTTP), File Transfer Protocol (FTP) and SMTP to mention a few. The rest of the high-level protocols are packaged in the commonly known as TCP/IP suite. PC users using analog MODEM connections utilize the Serial Line Internet Protocol (SLIP) to connect to the internet or sometimes use the Point-Point Protocol (PPP). These protocols are designed to encapsulate the IP packets in that they can be sent over a dialup connection to a given access provider’s MODEM. Related to TCP/IP; i.e, higher-level application protocols include; User Datagram Protocol (UDP) which can be used in place of TCP/ IP in special cases, other protocols in this scope include; Internet Control Message Protocol (ICMP), the Interior Gateway Protocol (IGP), the EGP (Exterior Gateway Protocol) and the BGP (Broader Gateway Protocol) [7].
TCP/IP Suite: Two-layered System
Figure 1: TCP/IP suite
The Internet protocol suite culminated from a research from the Defense Advanced Research Projects; DARPA building on the pioneering ARPNET. This was intended for secure communication between the security agencies for secure channels of communication share information between the respective agencies and stakeholders involved end-end. As discussed, TCP/IP is an end-end communication protocol which has evolved over time due to diverse technologies changes and needs, for instance, firewalls, address translators, web content cache have resulted into a significant change in advancement of TCP/IP suite [8]. This research is tailored towards identifying the current challenges in TCP/IP suite to position of a new design model built on its original principles to respond to the contemporary problems in the internet. Currently, TCP/ IP is summarized under the OSI model such that the application layer in the TCP/ IP entails application layer, presentation layer and session layers to the corresponding to the OSI model. The link layer on the other hand entails data link and the physical layers in the OSI counterpart.
Despite the significant changes that have been experienced on the internet space, the growing need for connections on the general cyberspace have presented new challenges in the dimensions of capacity bottlenecks that are currently affecting the general performance of TCP/IP-based connections. This research is tailored towards the theoretical demystifications of TCP/IP suite and how it operates to identify the source of the capacity bottlenecks. Qualitative approach of design is extensively deployed in this research to discuss and identify the capacity bottlenecks in TCP/ IP network connections.
Encapsulation is key in TCP/IP connection for hiding of protocols and services, it is usually affiliated with the division of the protocol suite into the general-purpose layers, in this sense, data packets are sent to their respective destination via the suite layers where they are encapsulated at every layer they go through. This technique is geared towards isolating upper layers from the details of bit transmission, e.g, Ethernet and collision detection mechanism whilst the lower level averts the need to know the details of each application and its respective protocols [9]. Whereas TCP/ IP is was revolutional at the time and has maintained the status to be used as the base internet communication protocol, the TCP/IP-based connections are majorly characterized by capacity challenges, in this sense, it has been established that TCP/IP suffer bandwidth and general performance issues emanating from a high number of connections in the related network. Capacity can be described as the number of connections that can be supported without the network having to be strained to affect its performance. A bottleneck in this perspective is a given point in the enterprise network where the flow of data is impaired, bottlenecks emanate from diverse causes including lack of ample capacity to handle the volume of data or where there is contention for internal resources, as a result, the flow of data is reduced in speed and capacity sometimes resulting into packet losses.
Capacity bottlenecks emanate from slow channel speed/ contention and excessive workload levied on the TCP/IP connection.
TCP/IP Communication Techniques
TCP/ IP suite is also known for its speed and contention problems, TCP/ IP speed is affected by different internal and external factors that tend to slow down the general speed of a TCP connection. Contention is one of the TCP/ IP bottlenecks in which there are more than one packets/ processes competing for a resource, in TCP/ IP scope, contention is evident in the connection-based TCP/ IP nature. In this essence, a connection has to be established for the communication to take place, contention could culminate into capacity problems such that only a limited number of packets can be sent at a time thus limiting the overall performance of a network in terms of the workload that can be supported at a time [10]. Contention may have been designed as a method of enforcing congestion window control, in the growing cyberspace, it is certainly becoming a real challenge top deal with this feature in TCP/ IP connections owing to the capacity issues it presents. The increasing cyberspace and the need for data communication needs a more accommodative protocol or an improved TCP/ IP in response to this problem as well as workload and latency issues to enforce a high-capacity TCP/ IP connection which is able to cater for the business needs of the current cyberspace.
Workload challenge resulting into capacity bottlenecks is a common feature on the internet owing to the bandwidth problems against the growing size and number of users on a given network. In this case, the initial bandwidth and resources are outweighed by the workload. In such a case, the resources strain hence resulting into a bottleneck that may not necessarily result into packet loss although still is a concern, in most cases, the bottleneck culminates into incredibly slow speed in the respective internet connection. This is a result of a queue of processes queuing for the respective resources to facilitate the communication, excessive workload is certainly set to reduce the overall performance of TCP/ IP connection, streaming workloads in TCP/ IP move large amounts of data from endpoints involved the common streaming workloads include; file transfer, backup or restore workloads or bulk data transfer, the main aspect of interest in the workloads is the network bandwidth.
Close to the topic, bandwidth can be defined as the values indicating the max capacity of a given TCP/ IP network to transmit data over a given network, usually, an optimal network has the highest bandwidth to eliminate bottlenecks as discussed in this paper. In this case, bandwidth is examined from the workload perspective, wired or wireless networks come with pre-defined calibrations illustrating the capacity that they can support [11], an excess to this designated value can result into poor performance of a network system. In some cases, however, the workload may be within range but still challenges in TCP/ IP communication as attributed to external factors such as Near-end crosstalks and far-end crosstalks in wired TCP/ IP networks. For a change and better performance of the underlying TCP network, alternate cables and configuration methods to eliminate the crosstalks is critical in improving the performance of networks in this perspective.
TCP/IP Application Layer Protocols
Also referred to as lag, latency in a network is the duration taken for a packet to be transferred from the source to destination measured in milliseconds. Latency tools measure the amount of time a packet takes as it is transmitted. An acceptable latency in TCP/ IP will depend on different aspects such as bandwidth and the overall configuration, however, the in-built architecture of TCP protocol may significantly increase latency thus affecting the overall performance of the network. In this respect, TCP congestion window is an internal structure designed to limit the number of data packets sent before an acknowledgement (ACK) is received. This is the main cause for latency problems in TCP communication as the packets need to be acknowledged before the next stream of packets can be sent in the network. If the ACK response delays, then it means that the next data packets will not be sent hence the capacity bottleneck [12]. Assuming that there is no packet loss, the sender sends the initial quota of packets corresponding to the congestion window, when the ACK packets are received, there is an increase in the TCP congestion thus an increase in the number of packets increases consequently, in the capacity scope, the number of packets is determined by the size of the congestion windows and the ACK packets. When there is high latency, the sender spends more time idle (low capacity) hence a reduced capacity to send data packets over a TCP channel.
Figure 2: TCP throughput vs latency
As illustrated in the above network performance measure, a higher throughput means there is low latency and consequently a high capacity in terms of the network performance/ capacity, the opposite means low throughput and thus high latency or low capacity.
The to-be TCP/ IP model is anchored on the challenges brought about by congestion windows which has been established to be the major cause for capacity challenges. Congestion window is intended to minimize packet loss hence its use, the focus in this model is thus focused on the ACK and bandwidth problem. The new model of TCP/ IP suite is set to contain additional protocols in its respective layers including;
The Stream Control Transmission Protocol as proposed in TCP/ IP model to address the capacity challenges is tailored to handle the end-end connection with less latency and RTT durations by ensuring packets get to the intended destination in the shortest period possible.
This is in nature is an adaptive bandwidth control mechanism in the sense that the protocol is built to handle streams of data of all sizes from the stream functionality [13]. The adaptive nature of SCTP is also extended to be implemented in the congestion window as defined by the multi-homing aspects of SCTP within TCP/ IP suite. In this case, the congestion window is made significantly sizeable in response to the low RTT and latency durations as facilitated by SCTP addressing and unicast for multiple properties described.
Figure 3: TCP/ IP suite design
The defining features of SCTP is that it is multi-homing and being a unicast with numerous properties. Confirming the TCP/ IP standard of being point-point, SCTP operates based on connection, additionally to this feature, SCTP supports a unicast with multiple properties/ paths. This is meant to eliminate the capacity problem emanating from the conventional unicast packet conveyance that usually results into slow ACK packets and consequently smaller congestion windows/ capacity bottlenecks. The addition of SCTP in the Transport layer will give TCP a new phase in which the data packets are communicated over multiple paths for faster ACK packets to be received and consequently an improved capacity as compared to the conventional TCP/ IP suite. This feature is generally termed as multi-homing as described in SCTP.
Figure 4: Addressing scheme
Unlike the conventional TCP/ IP model whose addressing ends at the IP address, the proposed model is set to include IP specifics such as ports just to ensure the data packets are acknowledged in the shortest period possible for an increased capacity as intended. Attributed to the multi-homing feature, the addressing feature also caters for address resolution issues with the provided port address along with the destined IP address as identified in the TCP headers.
Conclusion
Networks are the basis for digital communication, whereas there may not have been a problem in the past century regarding network performance, there is a certainly an issue in the current digital space in terms of the exponential size of cyberspace. The number of internet users, devices and data to be sent is rapidly increasing presenting new challenges of capacity, as presented in this paper, TCP/ IP re-modeling as described is crucial in responding to the changing needs of communication in terms of bandwidth and capacity issues.
Reference List
[1]. J.Sadowski,. Cyberspace and cityscapes: On the emergence of platform urbanism. Urban Geography, 41(3), 2020, pp.448-452.
[2]. C.Haynes, How to succeed in Cyberspace. Routledge, 2018.
[3]. A.M.Alotaibi, B.F.Alrashidi, S.Naz and Z.Parveen. Security issues in protocols of TCP/IP model at layers level. International Journal of Computer Networks and Communications Security, 5(5), 2017, p.96.
[4]. C.A.Grazia, N.Patriciello, T.Høiland-Jørgensen, M.Klapez, M.Casoni and J.Mangues-Bafalluy, Adapting TCP small queues for IEEE 802.11 networks. In 2018 IEEE 29th Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC) (pp. 1-6), 2018, September. IEEE.
[5]. A.Šljivo, D.Kerkhove, L.Tian, J.Famaey, A.Munteanu, I.Moerman, J.Hoebeke, and E.De Poorter. Performance evaluation of IEEE 802.11 ah networks with high-throughput bidirectional traffic. Sensors, 18(2), 2018, p.325.
[6]. K.Nguyen and H.Sekiya, TCP behavior on multi-gigabit IEEE 802.11 ad link. In 2020 International Conference on Green and Human Information Technology (ICGHIT) (pp. 58-61), 2020, February. IEEE.
[7]. S.Ponmaniraj, R.Rashmi and M.V.Anand. IDS based network security architecture with TCP/IP parameters using machine learning. In 2018 International Conference on Computing, Power and Communication Technologies (GUCON) (pp. 111-114), 2018, September IEEE.
[8]. L.Alhayali, T.Engineer, M.A.Mahmood, and A.S.Ahmed, A Proposed Study with the “DARPA Model” Network Issue Classifier.
[9]. S.Claypool, M.Claypool, J.Chung, and F.Li, Sharing but not Caring-Performance of TCP BBR and TCP CUBIC at the Network Bottleneck. In Proceedings of the Fifteenth Advanced International Conference on Telecommunications, Nice, France (Vol. 28), 2019, March.
[10]. B.Y.Kimura, D.C.Lima, L.A.Villas and A.A.Loureiro. Interpath contention in MultiPath TCP disjoint paths. IEEE/ACM Transactions on Networking, 27(4), 2019, pp.1387-1400.
[11]. W.K.Leong, Z.Wang and B.Leong, TCP congestion control beyond bandwidth-delay product for mobile cellular networks. In Proceedings of the 13th International Conference on emerging Networking EXperiments and Technologies, 2017, November.(pp. 167-179).
[12]. J.Luo, J.Jin and F.Shan. Standardization of low-latency TCP with explicit congestion notification: A survey. IEEE Internet Computing, 21(1), 2017, pp.48-55.
[13]. G.N.Vivekananda and A.Ilknur. A congestion avoidance mechanism in multimedia transmission over MANET using SCTP multi-streaming. Multimedia Tools and Applications, 79(23), 2020, pp.16823-16844.
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
