LM7 – Networking

Please read this entire page as few if any CISS 100 texts provide this additional information.  Note there is significant content in the optional resources below. If you are in the System & Network Administration program you should be reading this optional information as well.

Network Architecture

A network architecture is a conceptual blueprint that provides the necessary basis to define, build and maintain a physical and logical network and must provide and account for; 

        1. Fault Tolerance
        2. Quality of Service (QoS),
        3. Security
        4. Scalability.

(1) Fault tolerance allows a network to function normally even when components fail.  This is often accomplished through redundancy or duplication of equipment and media.

(2) QoS indicates the level of performance and services offered through a network.  We have observed the emergence and evolution of converged and unified communications (VOIP, Video/Web Conferencing, Multimedia Streaming, etc.) and these different technologies require different levels of service.   As an example, we do not want dropouts (i.e. lost or delayed packets) in our VOIP conversations but waiting 6 seconds instead of 5 seconds for a file download/transfer is probably acceptable.  In this capacity, QoS determines communications priority decisions in relation to the communications import to an organization.  Networks need a mechanism to manage congested network traffic and QoS can provide a mechanism to mitigate this complexity and provide satisfactory communications.

(3) Security is a critical component of networking since users must have confidence in communication services.  How many times have you heard me state that we think about security first in everything we do.  A network must (1) ensure confidentiality by restricting message access to intended and authorized recipients, processes, and devices.  The network must also (2) provide communications integrity by ensuring that information is available and that it has not been accidentally or intentionally altered during transmission (i.e. Information Assurance).

(4) Scalability determines a network’s ability to grow and react to future changes.  This includes both internal growth and external growth by joining other networks (e.g. Inter-networking).  We will see that both the OSI Reference Model and TCP/IP Protocol Model satisfy these core requirements.

Networking Elements

To properly understand network architecture and infrastructure (and keep in mind the Internet is a network) it is necessary to understand the 4 network elements comprised of;

        1. Communication devices,
        2. Messages and their encodings
        3. Mediums
        4. Protocols.

For completeness,  it also necessary to understand (5) network topologies

(1) Communication Devices:

We will leave the detailed understanding of the communication devices almost solely to the System and Networking Administration program however as a basis, many communications devices may be involved in addition to the hosts or end-devices (attached with Network Interface Card – NIC) participating in the communications.  These intermediary devices include: (a) network adapters/modems that connect users to networks (e.g. cable, telephone) (b) network access devices to connect end-users to the network (e.g. hubs, switches, WAP), (c) internetwork devices that connect networks (e.g. routers), (d) communication servers that perform routing services (e.g. DHCP, IPTV, Wireless broadband, etc.),  and (e) security devices (e.g. firewalls).  It must be noted these intermediary devices run a multitude of processes that route, regenerate, and retransmit data signals and manage the network, its security, and QoS.

(2) Messages & Encodings: 

Messages are the units of information that travel between devices (e.g. segmented Web pages, e-mail, Voice over IP, etc.).   Message encodings have several purposes that include security (encryption) and timing.  To understand the need for timing note it is well known that it is impossible to synchronize 2 computers and their clock cycles but it is possible to embed the timing in the signal as Manchester encoding does.  Rather than signal a 1 with a high state and a 0 with a low state (or vice versa), Manchester encoding uses a mid-cycle transition from high to low and low to high to encode the 1’s and 0’s.  The following illustration is from Wikipedia’s entry on Manchester Encoding.

(3)  Mediums:

As previously introduced, historically, interconnections have been a limiting factor in computing power and speed.  The physical layer network media is the foundation of any network and provides the framework for the network architecture.  Choosing and implementing the correct and appropriate type of connection medium is critical to data communications and therefore an organization’s success. The 3 mediums are: Fiber Optic, Copper Cable, and Wireless.

The messages may travel over several different mediums each with their own specific characteristics (fault tolerance, QoS, security, scalability, analog or digital, etc.) and protocols (e.g. Internet’s TCP/IP, Ethernet’s CSMA/CD, etc.). To understand networking mediums please read “Physical Layer Media” in the sub-menu structure.

(4) Protocols:

Protocols establish consistent rules for exchanging data between applications, services, and layers providing the foundation for communications congruence.  Protocols must necessarily determine: (a) the format of the message, (b) the data segmentation, (c) how intermediary devices share information about paths, (e) the methods to handle update messages between intermediary devices, and (e) the process to initiate and terminate communications between hosts. To understand protocols further, please read “TCP & UDP” in the sub-menu.

(5) Network Topologies

Lastly the 5th item necessary for proper understanding: Network Topologies although this is not a network element.  Network Topologies are covered in the textbook but should you want a deeper understanding please read “Network Topologies” in the sub-menu.  Every CIS student should understand Ethernet technology so if you have questions following the lecture please read about Ethernet & CSMA/CD in the sub-menu.  System and Network students should also read Wide Area Networking in the sub-menu.

Ok, so we established the foundational concepts above however we need a few more items for our basic understanding before we proceed.  As an introduction and overview of networking architectures, please read the “OSI & TCP/IP Models” found in this Networking and Internet sub-menu structure.  Also, please review and understand the content below.

Analog vs. Digital

Analog is continuous – http://en.wikipedia.org/wiki/Analog_signal

This continuous nature gives ultimate precision but detecting different states is complex and error-prone (e.g. what happens to waves if there is insufficient power or variances in the generation or reading mechanisms).

Digital is discrete – http://en.wikipedia.org/wiki/Digital_signal

Digital has fixed intervals (in computing interpreted as, 1/0 or on/off or true/false).

Circuit vs. Packet Switching:

Circuit switching maintains the connection path for the duration of the communication (e.g. landline phones).

Packet switching routes packets dynamically across data networks.

Here is a nice description of circuit and packet switching.

Now to get everyone thinking, we assess everything for security, does the single circuit switching path make it easier to eavesdrop/intercept communications (in contrast to packet switching) since all communication occurs over this single path.

Synchronous vs. Asynchronous Transmission vs. Communication:

Many texts present sync vs. async communications from the end-user or computer-mediated communications standpoint (i.e. synchronous phone call vs. asynchronous email) which is not incorrect (i.e. a phone call is synchronous and email is asynchronous when discussed at that level), but it is important to understand the difference at a more finely granular networking level.

Synchronous transmission requires connection management so that the parties know how (predetermined format) and when the transmission is occurring (connection-oriented). An example would be Web Conferencing. Note this does not mean that an end station knows precisely when a message is arriving only that it is arriving. This can occur at different levels (datalink, TCP, etc.). (Note our present text presents this concept in a misleading way as it seems to imply the end stations know when the message is arriving but it is impossible to know precisely when a synchronous communication will arrive when sent over a network).

Asynchronous transmissions, data is transmitted when it is ready to send, and the receiver must determine the timing inherent in or built into the signal.

Here is a more detailed description from InetDaemon.com.

Connection-Oriented vs. Connectionless

Here is a nice presentation on Connection vs Connectionless communications with excerpts below. Please see the TCP/IP resources in the sub-menu for a more complete presentation on the TCP 3-way connection setup handshake.

Connection-oriented means that when devices communicate, they perform handshaking to set up an end-to-end connection.  TCP/IP  uses a 3-way Handshake and the submenu is a complete writeup of TCP/IP but here is a nice succinct description: http://www.slashroot.in/understanding-tcp-three-way-handshake

Connectionless means that no effort is made to set up a dedicated end-to-end connection.

Simplex, Half-Duplex & Full Duplex:

Simplex – one direction only: http://en.wikipedia.org/wiki/Simplex_communication

Half-Duplex – one direction at a time (e.g. walkie-talkie)

Full Duplex – both directions concurrently (phone): http://en.wikipedia.org/wiki/Duplex_(telecommunications) (link also contains multiplexing description)

Now which one requires the most protocols?  Think about the phone, what would happen if you didn’t provide courtesy (protocol) and just continually spoke interrupting the other speaker?

Multiplexing

Recall we introduced Multiplexing in Operating Systems and defined it as the creation of several logical resources from a single physical resource.  In OS we further discriminated between space division and time-division multiplexing. Networking multiplexing refers to the ability to transmit several signals over the same channel simultaneously and this can be done using time-division multiplexing (e.g. packet switching) and frequency division multiplexing where different frequencies can be placed on a medium concurrently (e.g. DSL & WiFi).

Bandwidth

Bandwidth is a measure of the difference between the lowest and highest frequencies a media can transmit and is expressed in hertz (Hz).  Higher frequencies can transmit more data in a given period of time since they accommodate more transitions.

Throughput

Throughput is defined to be the amount of effective data that a network can accommodate during a given time and is usually measured in megabits per second (Mbps). Throughput is determined by the physical nature of the media, the network’s physical and logical configuration, and the network’s management protocols.

Wireless

To provide a foundational understanding please read about Microwaves that range from .3 GHz (300 MHz) to 300 GHz: http://en.wikipedia.org/wiki/Microwave.

Business-IT-Society Analysis

In addition to security, we assess everything with respect to the Business-IT-Society triangle and what is rarely discussed about wireless in CIS texts is: are these microwaves affecting us from a health standpoint?   A significant portion of medicine is based on statistical evidence.    We have only been immersed in environmental microwaves for 20 years and the present environment may even be saturated with microwaves.   Now consider the accelerating increase and evolution of both carrier waves and bandwidth (e.g. increase in the number of wireless signals and evolution of technologies bandwidth such as 1G -> 2G -> 3G -> 4G, etc., Bluetooth is now 4.0, Wireless is now 802.11x, etc.). We may not know,  detect, or diagnose the effects of these microwaves on our bodies for many years. 2009 Senate hearings on health risks of cell-phone microwaves/emf.

WANs

Every CIS student should understand the Internet so please read “Wide Area Networking & Internet” in the sub-menu.  For your reference, note that Wide Area Networking also presents multiplexing, circuit switching, and packet switching.

IPV4 & IPV6

Now we should also be aware that IPV6 has arrived and facilitates the IoT due to its increased address space: http://en.wikipedia.org/wiki/IPv6

IPV4 used 32-bit addressing (4 bytes) e.g. 187.56.201.125

IPv6 uses a 128-bit address, allowing 2128, or approximately 3.4×1038 addresses (facilitating IoT among other things

Physical/Logical Address Resolution

At some point, the system must resolve the logical to physical address mapping. TCP/IP does this with the Address Resolution Protocol (ARP).

Lecture Recording Pt 1

Lecture Pt 2

Lecture Pt 2

Additional Networking Resources/Labs

YouTube also has some good content and here is an example: https://www.youtube.com/@PowerCertAnimatedVideos