Internet Connection Types

Internet Connection Types and Modems

• The Internet is a global network of networks.
• The core of the Internet comprises high bandwidth fiber optic links that connect Internet exchange points (IXPs).
• Telecommunications companies and academic institutions primarily create these trunk links and IXPs.
• Internet service providers (ISPs) establish high-speed links within the datacenter that supports any given IXP.
• They use transit and peering arrangements to carry traffic to and from parts of the Internet they do not own.
• A tiered hierarchy of ISPs reflects their dependence on transit arrangements with other ISPs.
• Customers connect to the Internet through an ISP's network via that network's nearest point of presence (PoP), like a local telephone exchange.
• An Internet connection type is defined by the media, hardware, and protocols linking a local network at a home or small office to the ISP's PoP.
• This WAN interface typically follows a point-to-point configuration.
• Only two devices connect to the media, unlike Ethernet configurations.
• Digital modems typically establish connections to WAN interfaces, unlike Ethernet connections made with NICs and switches.
• Modems establish the physical link to the WAN interface and establish a method for identifying each network and forwarding data.
• This function is performed by a router that implements the Internet Protocol (IP).

Digital Subscriber Line (DSL) Modems

• Many internet connections use the public switched telephone network (PSTN), which is a national and global telecommunications network.
• The core of the PSTN is fiber optic, but its edge often still consists of legacy two-pair copper cabling, which is referred to as the plain old telephone system (POTS), local loop, or last mile.
• Digital subscriber line (DSL) uses higher frequencies available in copper telephone lines.
• Advanced modulation and echo canceling enable high bandwidth, full-duplex transmissions.
• There are asymmetrical (ADSL) and symmetrical types of DSL available.
• Asymmetrical DSL (ADSL) has a fast downlink and slow uplink, with the latest version (ADSL2+) offering download rates up to 24 Mbps and upload rates of 1.25 Mbps or 2.5 Mbps
• Symmetrical DSL versions offer the same upload and download speeds, making them useful for businesses and branch office links which require greater upstream data transfer.
• The customer network connects to telephone cabling through a DSL modem, either as a separate device or embedded into a SOHO router.
• RJ11 WAN port connects a standalone DSL modem to the phone point.
• RJ45 interface connects the modem to the router, while a filter (splitter) installed in each phone socket separates voice and data signals.
• Modern sockets tend to have built-in splitters and these can be self-installed.

Cable Modems

• A cable Internet connection is generally part of a cable access TV (CATV) service, where the CATV network is described as hybrid fiber coax (HFC).
• HFC combines a fiber optic core network with copper coaxial cable links to customer equipment or is described as broadband cable or just cable.
• Cable based on Data Over Cable Service Interface Specification (DOCSIS) supports downlink speeds of up to 38 Mbps in North America or 50 Mbps in Europe, and uplinks up to 27 Mbps.
• DOCSIS version 3 enables multiplexed channels to achieve higher bandwidth and installation of a cable modem follows the same principles as with DSL modem.
• The cable modem interfaces with the local router through an RJ45 port along with the access provider's network through a short coax using threaded F-type connectors.
• Coax links all premises in a street to a cable modem termination system (CMTS), which forwards data via a fiber backbone to the ISP's point of presence and onward to the internet.

Fiber Optic Connections (FTTx)

• The major impediment to LAN-like internet access is last-mile bandwidth, where copper wiring is often low grade.
• Projects that upgrade wiring to utilize fiber optic links fall under the umbrella term Fiber to the X (FTTx).
• Fiber to the Curb (FTTC) maintains some copper wiring to customer premises while extending fiber from the point of presence to a communications cabinet for multiple subscribers.
• Utilizing Very High-Speed DSL (VDSL) to support FTTC is common among service providers rooted in telephone networks, where VDSL achieves higher bitrates but at the cost of range.
• VDSL allows for both symmetrical and asymmetrical modes.
• Fiber to the Premises (FTTP) runs the service provider's fiber optic cable to the customer's building.
• This full fiber connection is implemented as a passive optical network (PON) with a single fiber cable is run from the point of presence to an optical line terminal (OLT) in a street cabinet.
• From the OLT, splitters direct subscriber traffic to an optical network terminal (ONT) installed in the customer's premises over a shorter fiber length.
• The ONT converts the optical signal to an electrical signal and then connects to the customer's router using an RJ45 copper wire patch cord.

Fixed Wireless Internet Access

• Wired broadband access is limited in rural areas or older developments with problematic cable runs for DSL or full fiber.
• A satellite-based microwave radio system, Geostationary Orbital Satellite Internett Access, provides larger coverage areas than other tech.
• Transfer rates depend on the provider and access package, but typically range from 2-6 Mbps up and 30 Mbps down.
• A drawback to geostationary orbit satellites is increased latency due to the signal travel of thousands of miles more than terrestrial alternatives.
• To create a satellite internet connection, the ISP installs a very small aperture terminal (VSAT) satellite dish aligned with the orbital satellite.
• Satellites are in high geostationary orbit above the equator, causing dishes in the northern hemisphere to point south and realignment is unnecessary.
• Coaxial cabling connects the antenna to a Digital Video Broadcast Satellite (DVB-S) modem.

Cellular Radio

• Cellular radio wireless networking facilitates communication over much larger distances (2.4 GHz and 5 GHz bands have limited range).
• Cellular networking is used by some IoT devices, while cellular digital communications standards are described as belonging to a particular generation.

Routers

• Current devices enable physical links and identifier a host's hardware interface.
• Ethernet switches and Wi-Fi access points forward frames using MAC addresses.
• Hosts can send frames to one another using MAC addresses on a network segement.
• Digital modems, ONTs, and cellular radios transmit data over DSL, cable, fiber, satellite, and cellular links to an ISP and this method does not require unique interface addressing (point-to-point link).
• A protocol is needed to distinguish between the private LAN and public WAN and an intermediate system with interfaces in both networks to connect a local network to the Internet.
• The Internet Protocol (IP) and a router function as the protocol and intermediate system.
• Routers forward packets around an internetwork using IP addresses, where a switch forwards frames using MAC HW addresses.
• An IP address contains the identity of both the network and a single host within that network, while a MAC address only identifies a hardware port.
• A SOHO router often routes between its local network interface and its WAN/Internet interface, while enterprise networks use different router models to carry out routing tasks.
• A LAN router divides a physical network into multiple logical subnetworks, where each becomes a separate broadcast domain and performance is reduced.
• Traffic subject to filtering rules are a security benefit of using a LAN router, which only has Ethernet interfaces
• A WAN or border router forwards traffic to and from the Internet or over a private WAN link, and has an Ethernet interface and a digital modem interface for the WAN.

Firewalls

• Controlling allowed computer connections and the types of accepted traffic is necessary after joining public and private networks with a router.
• A network firewall filters hosts and protocols; configured with a network access control list (ACL) that lists source/destination network addresses, protocol types, and allow/block status.
• These are deployable in private networks, where most routers implement some degree of firewall functionality.
• An appliance can act as a stand-alone firewall with deeper analysis of application protocol data and uses rules to determine what traffic is allowed,
• Unified threat management (UTM) appliances often implement firewalls to perform multiple other security functions.

TCP/IP Models

• A protocol is a set of rules that allows networked hosts to communicate data in a structured format.
• Protocols used together are designed to work as a protocol suite or the Transmission Control Protocol/Internet Protocol (TCP/IP) suite
• Network functions can be better understood by dividing network functions into layers, where lower layers encapsulate data from higher protocols.
• Each protocol then adds it's own header fields to data being transported from upper-layer protocol.
• The TCP/IP suite uses a with four distinct layers: Application, Transport, Internet, and LinK/Network Interface.

IPv4 Addressing

• The core protocol in TCP/IP is the Internet Protocol (IP), which provides network and host addressing and packet forwarding.
• An IP packet adds headers to whatever transport/application layer data its carrying to its payload, with two important header fields being the source and destination IP address.
• The two versions of IP are IPv4 and IPv6, with IPv4 being 32 bits long represented in its raw form as 11000000101010000000000 000000001.
• The 32 bits can be divided into four groups of eight bits, known as "octets."
• Humans use dotted decimal notation, where each octet is represented by a decimal value (0-255) separated by a period and the binary 11000000 10101000 00000000 00000001 would be 192.168.0.1.
• Every IPv4 address has a network number common to all hosts in the same IP network (network ID) and a host number (host ID) that identify a host in a IP network.
• These are distinguished by combining the address and a network prefix, represented by a 32 bit value with a given number of contiguous bits all set to 1.
• A prefix with 24 bits represents the binary value, 11111111 11111111 11111111 00000000 or is expressed in slash notation as /24 (255.255.255.0).
• When combined with an IP address, the prefix masks the host ID to reveal the network ID, where a binary 1 in the prefix represents a binary digit or where the IP address is a part of the network ID.
• Packets travelling via IPv4 are sent through a process used to determine which subnet to use.
• The protocol compares the source and destination IP addresses with the host's subnet mask, if the masked portions match, the interface of the destination is on the same IP network, where most hosts are configured with a default gateway parameter.
• The default gateway is the IP address of a router interface that the host uses to forward packets to other networks, so it must be a part of the same IP network.

Public and Private IP adresses

• To communicate on the Internet, a host needs a unique public IP address, assigned by ISPs, but few companies have sufficient public IPv4 addresses.
• Private addresses allow multiple devices to be used behind a single public IPv4 address on the World Wide Web.
• Private address ranges (not routable on the Internet) are:
  • Class A: 10.0.0.0 to 10.255.255.255
  • Class B: 172.16.0.0 to 172.31.255.255
  • Class C: 192.168.0.0 to 192.168.255.255
• Address classes and Default Subnet Masks: derive from earliest form of IP, IP did not include the concept of subnet masks and hosts would identify the network ID with the class.
• Subnet masks mirror functionality and are referred to as "default masks."
• Hosts configured with a private address cannot access the internet directly and will need a mechanism to allow it through (i.e Network Address Translation)
• Methods to get internet access while using private addressing are:
  • Routers configured with a valid public IP to translate from private and public.
  • A proxy server that fulfills requests for Internet resources on behalf of clients.

IPv4 and IPv6 Host Address Configuration

• Every host must have a minimum IP address and subnet mask when configured to use an IPv4 network.
• Hosts should be configured to have full network use and to properly use a modern internet.
• IPv4 addresses and subnet masks can set up manually, where the address is entered as decimal numbers separated by periods (i.e 192.168.0.100).
• An IPv4 subnet mask is entered in dotted decimal notation, similar to 255.255.255.0.
• The mask can be represented with the mask length in bits and is used with an IP address such as 192.168.0.100 identifies 192.168.0.0 as the network ID.
• Besides the address and mask, to make the host fully functional two other things need to be configured.
  • One is the Default Gateway , a router's IP address, and a required setting to communicate locally, although it isn't compulsary.
  • Second is the Domain Name System (DNS) server IP address, which are essential for locating resources on the Internet and provide resolution of host and domain names to the IP addresses.
• The primary DNS server address can also be commonly configured as the same as the gateway.
• Routers forward DNS queries to a secure resolver, and often dual DNS server addresses (preferred and alternate) to provide redundancy.

Static versus Dynamic Host Address Configuration

• Static addressing requires manual configuration that is time consuming and error prone, or an administrator when a host moves to a new IP network.
• Static addresses should be configured for router interfaces and server applications that need a fixed IP address.

IPv6 Addressing

• The available IPv4 public addresses are limited compared to internet-connectd devices so IP version 6 (IPv6) is intended to be a replacement.
• The IPv6 address is a 128-bit number and can have exponentially more address values when compared to the 32-bit IPv4 addresses.
• IPv6 addresses are written in hexadecimal notation, where a hex diget can represent a four-bit binary value (nibble).
• To express a 128-bit IPv6 address in hex, the binary address is divided eight double-byte (16-bit) values delimited by colons.
• To shorten typing in configuration for double byte addresses with leading zeros, the bytes can be ignored (example 2001:0db8:0000:0000:0abc:0000:def0:1234).
• One contiguous series of zeroes can be replaced with a double colon marker to become 2001:db8::abc:0:def0:1234
• An IPv6 split the address into a network ID (first 64 bits) and the second 64 bits which are the specific interface, while subnet mask is not needed due to a fixed size.
• Using prefix notation is used to denote network addresses with /nn, where nn represents the length of the prefix in bits (within the 64-bit network ID)
• IPv4 addresses commonly have single IP addresses per interface, however IvP6 interfaces are configured with multiple.
• The main types are global, unique to the internet (equivalent to to public IPv4 addresses as link-local for communicating with neighbor hosts.
• Links-local addresses start with fe80:: while global begin with a 2 or 3.
• Most hosts obtain global and link-local addresses from the local router, a process referred to as StateLess Addressed Auto Configuration (SLAAC) and do not need to be configured with a default gateway.
• Router discovery in IPv6 is completed with the Neighbor Discovery (ND) protocol, which allows for stateless address configuration and also allows a host to discover a router as well as perform interface address querying functions (used by ARP in IPv4).

Transport Protocols and Ports

• Network hardware and protocols that have been covered focus primarly on moving frames and packets.
• At the link layer, Ethernet allows users to send one another frames with MAC addresses.
• IP provides addressing and routing functionaility for a network of networks such as and ethernet.
• Then at the transport layer it provides a way to identify each type of network application.
• It does this with each type of network application by applying port numbers between 0 and 65535 (i.e HTTP port 80, E-mail trasmission = port 25)
• Application segments are multiplexed using the port numbers to the same network link.
• The two protocols to determine port assignment is Transmission Control Protocol (TCP) and User Datagram Protocol (UDP).
• TCP provides many mechanisms to overcome a lack of reliability -Connection-oriented via using SYN handshake
  • Trackable by assigned sequence number so packets don't fail. -Acknowledges reception as well as the ability to resend missing data. -TCP's multiple header fields requires multiple header fields with can increase each packet to 20 bytes. -TCP is used commonly when information can't be damaged, i.e HTTPS, Secure Shell (SSH).

Transmission Protocols Continued

• User Datagram Protocol (UDP) -Faster but less reliable -Connectionless -Non guaranteed sequencing - doesn't send sequenced and ack packets between end points. -There is no gurantee regarding message delivery sequence. -Used in time-sensitive aplications (i.e. Voice and video). Missing information can be lost as a "glitch". -The reduced overhead enables faster delivery, with an application layer that can be used to control delivery reliability.
• Dynamic Host Configuration Protocol (DHCP) -Used by clients when requesting IP information -Not often supported by TCP and must be UDP to broadcast transmissions. -Used to make sure response packets are resent after the process time out repeatedly.
• Trivial File Transfer Protocol (TFTP)
  • Devices using network use this protocol which is typically a configuration file.
  • Application protocol uses it's own acknowledgement messenger.

TCP & Ports assigned

• Servers uses Port numbers to enable to connecton (ex: file downloads, secure connections )Server port numbers are assigned by the Internet Assigned Numbers Authority (IANA)
• File Transfer Protocol (FTP) enables download of files across a network (port 20)-Data Connection
• File Transfer Protocol (FTP) enables download of files across a network (port 21)-Control Connection
• Secure Shell (SSH) is used on port 22 to command line connect to a server.
• Telnet is used on port 23 with connects to the command line of a server.
• Simple Mail Transfer Protocol (SMTP) uses port 25 when transfering messges across a network.
• Domain Name System (DNS)TCP/UDP enables facutlative identification with identifying addreses to host through ports 53's.
• Provides the IP addres on a dynamic hosts through clients (server port 67) with servers that are a dynamic configurations.
• HTTP Transfer Protocol clients also use port 68's for configuration and requesting.
• HTTP is applied of ports 80's enabling websites, services and the unsecure versions to connect.
• Ports 110's uses the mail messages of a server, with retrieving email messages using a legacy features, that have ports 137-139, with other network features.
• Ports 143's on TCP reads and manages email messagin on server mailboxes.
• Protocols that query status include : Simple Network Management Protocol (SNMP)-(161 UDP) ,the management of network reports(162 UDP), and network resources through lightweight directory access (389 TCP).

More Ports numbers

• HTTP enables secure web services on port 443 TCP
• Ports 445 provides compatilbity with windows with file and printed shared servers.
• 3389 ports on TCP supports secure connections.

Dynamic Host Configuration Protocol -DHCP

• Interfaces used to make static configurations, is sometimes mistaken, creating duplicate IP's that need manually updated for network changes.
• DHCP servers are used to avoid this and can allocate/change the ip addresses and subnet masks at any given time by any connecting host.

DHCP Cont...

• A scope, usually is defined as a range of addresses, clients may use ,and are unique. A LAN may have a scope.

DHCP leases

• To use a DHCP, TCP/IP configures to automatically accept the ip address to which then DHCPOFFER relays data to (such as information to the defaut gateway, IP's etc).

• A client may select if it is to broadcast packets into the network or the offer by which it receives packets. For this to work, the server must reply through DHCPOFFER packets and the client. -To check if there is a unused address for every message, the Client needs to check ARP data and the server has limited time that allows address requests and release

DHCP Reservations

• Hosts must utilize static- IP addresses if a dynamic IP addresses are not available. If managing routers are difficult the following steps will insure static address. - A DHCP server can assign a unique address for every static device with their associated mac adress, creating leases.

DNS (Domain Name Service)

• Binary uses IP addresses to locate Hosts thru configurations, to correctly find and remember host names thru configuration,
• Host Names are set when the OS is installed and host names are local to the network.
• To resolve conflict (name duplication) you can create a name combined with the host an a domain,creating FQDN. The .exe domain or host is a suffix (ex:"widget-.examplenut".)
  • DNS has a directory that has a name distribution.
• The root of directory contains, servers such as A to M, to connect at the TLDS.
• Tlds are managed thru icann.org thru countries and code that generally appointed by the government Clients must contact the host by using a address record to access the DNS server and server. The domain is looked within the local cache by looking at a"stub server". With tcp/ip having 53 port to which DNS is enabled from the server on clients.

How DNS Works

Clients uses (FQDN) to look for a host and if there is no lookup from the local cache. The local server will look throughout the pre- existing IPS of existing host.

DNS Continued

To transfer your clients quest by which the local server look throughout the host, the dns (IP info for looking up client domains over internet) records needs setup, acting with authority and needs to be assigned by the DNS (seperate to the client). The names have a zone for each server to resolve queries.Resource records can be updated manually via client servers over networks.VLAN's and networks are split into one unmanaged broadcasted (switch to all hosts) domain.

The problems that can happen on VLANS

That performance can drastically reduce placing in one domain switch thousands of hosts.To split this u, manageable switches assign groups by turning them into Vitural Lans.(VLAN)The switches configure the port interfaces with switches.Ports are assignes, if they do not connect , the networks use routers as IP routers or DCHP/ DNS server for use with Vitural LAN; each VLAN is a security zone.For isolation the devices use traffic filtering for security, which support data and voip. The hosts have virtual pirvate netwoks to access LAN'S in place or installing equipment there.The Hosts can connect with remote acccess and switch or AP(access point) with security enabled thru encyption. VPN and secure connnections enable proper authentication to connect to the remote host over internet (by which are restricted to bandwith at remote devices/locations)