[#netcable]: Cabling

Connectivity: Cabling/Wiring

See: Glossary entry for “cable”.

Copper cabling

This is commonly used for Ethernet communications.

Type/Style
Twisted Pair
[#cprtpcor]: Core type
Stranded
Lots of little fibers connected together. More capable of handling a break in a fiber, as the conductivity may just flow through some of the other fibers.
Solid Core
Used for in-wall wiring that gets stored out of the way (e.g. between walls). May provide superior conductivity, at the costs of being more fragile and may be a higher price.
[#tptwists]: Discussion on the twists

The twists may help prevent electricity from moving in the same physical direction for long enough to generate sufficient “cross-talk”-style intereference. More twists results in less likelihood of cross-talk, and so is higher quality.

However, the term “twist” refers to wires that are curved around each other. More twists means more copper is required, for the same reason that a heavily curved line is a less direct route than a barely-curved line (which is less direct than a completely straight line). Naturally, more copper will also involve more of the plastic insulation that is tightly wrapped around the copper (following the same twisted line as the copper). So, presumably since more physical material is required, more twists tends to be more expensive.

Wikipedia's article on RG-59 (which is about different writing, not twisted pair), does have the following commentary about unshielded twisted pair (“UTP”) category six (Cat6/CAT6) : “UTP such as CAT6 has too many twists in the twisted pairs which results in a poorly performing media for analog signals over long distances”.

[#catcblcp]: Twisted Pair Copper Cabling Categories

The copper cabling itself may fit one of several standards. Following are some standards for the copper cabling, and some similar/related standards. The term “Category” may often be abbreviated as “Cat.” with a trailing period) or, probably more commonly, as “Cat” (without a trailing period). Also, in the latter case, the space after the word “Cat” may often be omitted.

One thing that may help identify higher quality cable is having more twists.

Category 1
Very simple wiring, which may be used for extremely simple communications. For example, Cat 1 may be used for a doorbell which communicates when a button is pressed.
Category 3
Used by telephone communications. (A typical telephone wire may use CAT3 cabling with RJ11 connectors.)
Category 5

More commonly used is Cat5e (Category 5 enhanced). This is generally suitable for communications up to 100Mbps. Communications at 1000Mbps / 1Gbps might be possible, although may not work. (Such high speeds are even less likely to function if the wiring being used is actually Cat5 rather thatn Cat5e.)

The ends tend to be RJ-45 jacks using either the EIA/TIA T568A standard or the EIA/TIA T568B standard. Some wiring details may be in the section about crossover cables.

[#cablcat6]: Category 6 (and Category 6a, a newer standard)

Category 6a is also known as “Category 6A”, or Cat6a (or “Cat6A”).

Generally suitable for communications up to 1000 Mbps / 1 Gbps. 10gbps may be permitted with reduced range.

Wikipedia's article for “IEC 11801”, section for “Category 7” says, “in 2008 Category 6A was ratified and allows 10 Gbit/s Ethernet while still using the traditional 8P8C connector” (RJ45). “Therefore, all manufacturers of active equipment and network cards have chosen to support the 8P8C for their 10 Gigabit Ethernet products on copper” ... “These products therefore require a Class EA channel (Cat 6A).”

Wikipedia's article on “10-gigabit Ethernet”: section called “10GBASE-T” notes, “Although category 6a is required to reach the full 100 metres (330 ft), category 5e is good for up to 45 metres (148 ft) and category 6 will reach 55 metres (180 ft).” Wikipedia's article on “Category 6 cable”: “Maximum length” section states, “When used for 10GBASE-T, Cat 6 cable's maximum length is 55 meters (180 ft) in a favourable alien crosstalk environment, but only 37 meters (121 ft) in a hostile alien crosstalk environment, such as when many cables are bundled together.” Testing is recommended (by the text on Wikipedia) for installations that use Cat6 instead of using the newer Cat6a. (For more information about cabling distance limitations, see copper twisted pair length limits.) Wikipedia's article on “Category 6 cable” states, “Whereas Category 6 cable has a reduced maximum length when used for 10GBASE-T; Category 6a cable, or Augmented Category 6, is” ... “improved” ... “allowing 10GBASE-T to be run for the same distance as previous protocols.”

[#cablcat7]: Category 7 (and Category 7A, a newer standard)

This is not a widely-deployed category of cable, but is being mentioned here for some completeness (rather than skipping from Cat6 to Cat8). Wikipedia's article for ISO/IEC 11801, section on “Category 7” (e.g., version from February 2021) says, “Category 7 is not recognized by the TIA/EIA at the time of this writing.” Also, the following section, titled Category 7A, states, “As of 2017 there is no equipment that has connectors supporting the Class FA (Category 7A) channel.” “Category 7A is not recognized in TIA/EIA.” “As of 2017 there is no equipment that has connectors supporting the Class FA (Category 7A) channel.” Connectortips.com: “What’s new in infrastructure CAT-5 and CAT-8 balanced pair connectors?” (from April 22, 2016) says, “CAT-7a usage is much more common in Europe and Asia versus North America. Overall volume is still smaller than CAT-6a usage globally. ”

“Class FA (Class F Augmented) channels and Category 7A cables, introduced by ISO 11801 Edition 2 Amendment 2 (2010), are defined at frequencies up to 1000 MHz”... “IEEE 802.3bq working group ratified the amendment 3 which defines 25Gbase-T and 40gbase-T on Category 8 cabling specified to 2000 MHz. The Class FA therefore does not support 40G Ethernet.”

Well, not officially anyway. The standard was defined with the thought process being that 40 gigabit Ethernet 50Gbase-T might be feasible. What research demonstrationed is: “ Simulation results have shown that 40 gigabit Ethernet may be possible at 50 meters and 100 gigabit Ethernet at 15 meters.” “In 2007, researchers at Pennsylvania State University predicted that either 32 nm or 22 nm circuits would allow for 100 gigabit Ethernet at 100 meters.”

[#cablcat8]: Category 8

Wikipedia's article for ISO/IEC 11801, section on “Category 8” says, “is defined up to 2000 MHz and only for distances from 30 m to 36 m”. “Category 8 is designed only for data centers where distances between switches and servers are short. It is not intended for general office cabling.”

An earlier part of that article, in the section on Category 7A, notes, “the IEEE 802.3bq working group ratified the amendment 3 which defines 25Gbase-T and 40gbase-T on Category 8 cabling specified to 2000 MHz.” So Cat8 can support 40Gbase-T.

Wikipedia's article for ISO/IEC 11801, section on “Category 8” seems to identify two different types of Cat8 cable. The first, called “Class I” (that likely being a roman numeral one), a.k.a. “Category 8.1 cable”, is “fully backward compatible and interoperable with Class EA (Category 6A) using 8P8C connectors” (As noted elsewhere, 8P8C is often called RJ45.) “Class II”, a.k.a. “Category 8.2 cable”, is “interoperable with Class FA (Category 7A) using TERA or GG45.” (Those are different types of termination ends, different than 8P8C “RJ45” connectors.)

[#cablctnw]: Newer cabling standards

Wikipedia's article for “ISO/IEC 11801”: “Class F” section states, “The Category 7 cable standard has been created to allow 10 Gigabit Ethernet over 100 m of copper cabling (also, 10 Gbit/s Ethernet now is typically run on Cat 6A).” (Some hyperlinks have been removed from the quoted text.) There are also “Category 7a”, “Category 8”, “Category 8.1”, and “Category 8.2”. Wikipedia's article for “ISO/IEC 11801”: “Category 8” section says that Category 8.1 is “fully backward compatible and interoperable with Class EA (Category 6A) using 8P8C connectors”. (An example of such an “8P8C” connector, which identifies this as an 8 pin connector, is the connector commly named RJ45. This is discussed more by network connectors.) It is interesting that 8.1 refers to “6A”, not Cat7a. Much earlier, the page (Wikipedia's article for “ISO/IEC 11801” (last version that was updated in 2014) noted, “A major revision, Edition 3, is being prepared which will unify requirements for commercial, home and industrial networks.” The article's 18:04 update from January 23, 2018 said, “A major revision was released in November 2017, unifying requirements for commercial, home and industrial networks.”

Notes from Wikipedia's article for IEC 11801 in 2020 (March) state, “Category 8 is designed only for data centers where distances between switches and servers are short. It is not intended for general office cabling.” While some research suggested Category 7A might support 40G or 100G (possibly with range limitations of 15 meters or 100 meters), a newer standard “defines 25Gbase-T and 40gbase-T on Category 8 cabling specified to 2000 MHz. The Class” FA “therefore does not support 40G Ethernet.”

Newer CAT cables

What are the differences between Cat5, Cat6, and Cat7 Ethernet cables?” also mentions as low as Cat1 and as high as Cat8.

Wikipedia: IEC 11801: section called “Classes and categories” mentions various categories. Lower on that same page, further details are provided for some of the categories. The “Classes and categories” section also has hyperlinks to details on specific categories.

Coaxial cable (also known as “coax cable”, or “co-ax”)
Name

The copper core (wire in the center) is typically rather round, as is the insulation. Outside of the insulation may be yet more copper wiring, and then the outer layer of plastic. If the cable were straightly cut, both the wire and each of these additional layers will likely look rather round. The center can be called an “axis”. The “co-” syllable refers to the fact that each of these components shares the same common center axis (which is the center of the round wire).

RG-59

Capable of transmitting data for a small distance. Many times, the cabling does not need to be longer than three or four feet long, so the distance may not be a problem.

RG-6
This is superior to RG-59 (similar to what would be expected if the hyphen were treated like a decimal point).

Possibly related: Wikipedia's article on “RF connector” describes a “coaxial RF connector (radio frequency connector)”. Such an RF connector is likely what is supported by the “New NES” (Wikipedia's article for “New-Style NES” mentioned “an RF connection is the only way to connect the system to a television unless” one “uses an RF-to-RCA converter.” (Or, of course, if one modifies the console, which that sentence did note. Perhaps such modifications have been somewhat popular?)

Styles of cable protection (from EMI)

More advanced shielding may help to reduce “electrical/magnetic interference” (“EMI”). On the downsides, the shielding can add to expense, and may make the wiring less flexible/bendable and/or prone to break.

[#utp]: Unshielded Twisted Pair (“UTP”)

This type of cabling involves having a pair of wires. Actually, much more commonly, there are multiple pairs (perhaps four pairs, totalling 8 individual wires). (Each “wire” itself may be “stranded” consisting of smaller strings of copper.)

UTP lacks the “shield” in STP, which typically makes it cheaper and possibly a bit more flexible. However, UTP is more prone to experience errors from EFI (most notably “crosstalk”).

[#shieldtp]: Shielded Twisted Pair

A “shield” of cabling surrounds each pair of wires.

“STP” is an abbreviation that is often used for “shielded twisted pair”, although “STP” is also sometimes used to refer to variations such as “screened shielded twisted pair” which include a shield. (Perhaps this might also, hopefully less commonly, also refer to variations like “screened twisted pair” which might not have the sheilding around the individual pairs, but which may provide some defense against EFI that UTP lacks.) Also, the “Spanning Tree Protocol is often abbreviated as STP (and “Spanning Tree Protocol” is a subject used by many technicians who work with networks.) So, some care may be worthwhile when dealing with that specific abbreviation for network cabling.

[#scrndtp]: Screened Twisted Pair and others
“Screened twisted pair” may be referred to as “ScTP” or some other abbreviations. This is basically another layer of metal surrounding the wires to help reduce interference. Unlike STP cabling, which surrounds individual pairs of wire, this “screen” surrounds all of the pairs of wire. (There is also “screened shielded twisted pair” which has both the screen surrounding all pairs, and also shield(s) surrounding each individual pair of wires.) The screen may often be considered to be a sort of “foil”, and so an abbreviation like “F/TP” may also be used (instead of using “ScTP”). Both abbreviations have similarity to a standard communications protocol (File Transfer Protocol (FTP) or Stream Control Transmittion Protocol (SCTP). And, of course, “STP” has multiple meanings, as discussed in the section on shielded twisted pair.

In addition to the information just provided, other possible abbreviations for variations may exist: see Wikipedia's article on “Twisted pair”: section called “Cable shielding”.

Insulation
[#plenumcb"]: plenum (cabling)

The term “plenum” refers to space that exists between walls, or space above ceilings but below whatever is above the ceiling (the next floor, or perhaps the roof?), or below a floor (and typically above the next-lower ceiling, unless it is the bottom floor).

Plenum is also the name of a rating assigned to the outer coating of some Ethernet wire. Non-plenum wire would, generally, be cheaper.

Some fire code safety requirements may demand the use of plenum cabling. The reason this may matter, even if there is not a fire, is that non-plenum cabling can be dangerous. If the non-plenum cabling gets heated, it may start to melt. Even if it hasn't melted flammably or even visibly yet, this could cause the wire to give off some fumes that are dangerous. People breathing in the affected air may be negatively affected, but might not be able to smell the problem, and so may not notice right away what is happening.

A copy of Mike Meyer's All-In-One Net Plus book (Chapter 8) notes that plenum-rated cable “costs about three to five times as much as PVC-rated cable. Most city ordinances require the use of plenum cable for network installations. Bottom line? Get plenum!” (Don't you always love it when advice is given, and that advice involves needing to spend more money?)

Another safety tip: Make sure that any network cables do not touch, go directly over, or really go near flourescent lights.

Order of wires
Straight-through

See the section about network “patch” equipment (subsection called “patch cable”).

[#xovercab]: crossover

If two 10mbps Ethernet NICs are directly connected with a patch cable, they would often not be able to communicate at all. (This has been resolved with newer equipment: Wikipedia's page on “Medium Dependent Interface”: “Auto-MDIX” section notes that for infrastructure devices (such as a switch or router), “all 1 Gigabit or 10 Gigabit devices in practice” tend to support Auto-MDIX. There have been some claims of gigabit equipment that hasn't fixed this, but such a lack of MDIX support is quite rare with gigabit equipment, to the point that many technicians doubt the existence of such problems, and Wikipedia's article notes that “in [actual] practice”, it isn't an issue with Gigabit equipment.)

The reason this would not work is that both devices would try to transmit out wires 1 and 2, and tried to receive on wires 3 and 6. Since both devices were transmitting, and not listening, on the same wires, no effective communication occurred. The solution, by design, was for standard NICs (in computers) to connect to infrastructure devices such as an “ethernet hub”. A hub was designed to listen on wires 1 and 2.

(This is shown by Wikipedia: Ethernet_MDI_to_MDIX.svg / Wikipedia: 525px-Ethernet_MDI_to_MDIX.svg.png), referenced by Wikipedia: Medium-dependent interface. This shows the intentions of a regular computer using the “NIC MDI” electrical signals, and how that matches with the “Hub MDI-X” used by hubs that have different expectations.)

So, as long as all computers used the “NIC MDI” electrical expectations, and all of those computers connected to hubs designed to use the “Hub MDI-X” electrical expectations, everything matched nicely with standard straight-through cables, and this worked out okay.

When people needed to connect two computers together, and a hub wasn't convenient (because hubs are more expensive than cabling, or perhaps because hubs required an electrical power outlet), a solution was to just use a “crossover” cable which flipped the pair of wire number one and wire number two with the pair of wire number 3 and wire number 6.

Wikipedia: Ethernet MDI crossover.svg (as referenced by Wikipedia: Medium-dependent interface) shows the wiring: Each typical NIC transmits out on pins 1 and 2 (1 for positive, 2 for negative) and receives on pins 3 and 6 (3 for positive, 6 for negative). Another graphic: Crossover Adapter Wiring (JPG) / Crossover Adapter Wiring (JPG) found from Amazon: Cable Matters: 2-pack crossover adapter

Sometimes having a cross-over cable would cause the problem it was designed to solve: a computer may be trying to connect to a hub, and only a crossover cable was available. Some devices would have a physical switch to alter whether the connectivity of a single port generally the last, highest-numbered port) was treated like straight-through or crossover. (For example, the Netgear Ethernet Hub EN104TP was a 4 port hub. It had a switch labelled “Normal/Uplink” and this would alter the functionality of the fourth port. A photo of this device is shown at Wikipedia's photo of a 4 port Netgear Ethernet hub EN104TP.)

When using official colors recognized by the TIA/EIA T568 standards, this meant one end would use “TIA/EIA T568A” and the other would use “TIA/EIA T568B”.

There are numerous sites that show the preferred wiring order. One of the best is Wikipedia's article for “TIA/EIA 568”: “Wiring” section. The main trick (which can be easier when seeing the photograph) is to make sure that pin #8 is using brown, so that the wires aren't placed in the backwards order. When the clip is on the top of the cable's connector, if a person is standing on the end of the connector with the cable, brown needs to be on the left side and striped needs to be on the right side.

Some pictures showing the network wiring are available from numerous sources, including RJ-45 UTP Wiring Guide and ArchonMagnus.com guide and ProSoundWeb guide and more pictures related to crossover cabling.

Wikipedia's article for “Category 6 cable” notes, “T568B is a deprecated standard in the US and no longer supported by TIA.”

Wikibook: Network Plus Ceritfication: Wiring Standards: “Stright vs. crossover” notes, “1000BASET (Gigabit crossover), which uses all four pairs, requires the other two pairs (1 and 4) to be swapped and also requires the solid/striped within each of those two pairs to be swapped.” (The page also has a nice diagram for helping make a cable with that particular standard.)

A far less common, although nice, practice is to mark crossover cables with a non-transparent (e.g. black) “x”, with the marking located on either the connector or the wire. This might not be a heavily-recognized official standard, but can help people to quickly notice the fact that the cable is crossover (without needing to look at the wires).

There are “crossover adapaters”, which may be a small piece of plastic with a little bit of copper, which allow a person to plug in an Ethernet cable. The other end may be an Ethernet port (needing another cable) or a plug (which doesn't need another cable, but which may be part of the larger adapter and so may be less convenient when trying to plug into an Ethernet jack without abundant nearby space). Plugging in either a patch cable, or a crossover cable, will result in connectivity as if the other type of cable was used.

[#cprtplen]: Distance limitations (for twisted pairs of copper wiring)

While some types of (fiber-optic) cabling may provide reliable signals for distances that are multiple kilometers (2, or even 80) in length, standard copper twisted pair cabling may commonly have an official limitation closer to about 100 meters in length.

Wikipedia's article on “Category 6 cable”: “Maximum length” section notes, “This consists of 90 meters (300 ft) of solid "horizontal" cabling between the patch panel and the wall jack, plus 10 meters (33 ft) of stranded patch cable between each jack and the attached device. Since stranded cable has higher attenuation than solid cable, exceeding 10 metres of patch cabling will reduce the permissible length of horizontal cable.” Wikipdia's article on “Category 5 cable”: section called “Maximum cable segment length” cites a similar logic, noting the idea of “90 metres of fixed cabling, two connectors and two patch leads of 5 metres, one at each end.”

For some more notes related to distance, see: Category 6(a) cabling.

[#cprcblnd]: Copper Cable Ends

There are multiple:

RJ11

RJ-11 or RJ11 is not typically used for Ethernet, but has often been used for land-line telephone service with Cat3 cabling. DSL service has used it. Some other equipment, like the Adtran ActivReach NetVanta switches, may use it which could be helpful in old buildings where cable replacement may be prohibited or undesirable. However, running computer networking over such cabling would typically be considered notably inferior to at least Cat5e, and often viewed as sub-standard. The main issue with such a solution might be reliability compared to setups that are using higher quality cabling.

DocPlayer.net: Adtran notes, “he ADTRAN® innovative ActivReach™ technology, which provides a fresh approach, and removes these trouble-some constraints.” such as “Ethernet is limited to 100 meters (328 feet)” and “Approaches to delivering Ethernet over CAT3 are limited to 10 Mbps”

8P8C/RJ45

The phrase “8P8C” may be more accurate, although if you want to get that technical, 8P8C may also be more ambiguous because other connectors might use 8 pins/positions and 8 contacts.

Some further text is currently at: Glossry: R-terms: Registered Jack 45

GigaGate 45 (“GG45”) / Augmented RJ45 (“ARJ45”)

Connectortips.com: “What’s new in infrastructure CAT-5 and CAT-8 balanced pair connectors?” (from April 22, 2016) says, “The GG45 connector is not backward compatible with older RJ45 connectors as the contact pairs have been relocated to improve electrical transmission and crosstalk isolation.” Based on https://en.wikipedia.org/wiki/GG45#Details, it appears to have 12 connectors: 8 on one side (similar to typical connectors that are often called RJ45) but also 4 more.

TERA
Connectortips.com: “What’s new in infrastructure CAT-5 and CAT-8 balanced pair connectors?” (from April 22, 2016) discusses this a bit. The connectors are bigger, but may offer more usage flexibility which could make them more useful than the smaller GG45 (or RJ45) connectors.
[#netptceq]: network “patch” equipment
[#patchcab]: patch cable

A “straight-through” Ethernet cable. The term “patch cable” might most commonly refer to a fairly short (perhaps one meter/3 ft) cable, but is always referring to the idea of “straight-through” cabling. “Straight-through” simply refers to the wiring matching on the connectors on both sides: If one end is using EIA/TIA T568A then the other end also needs to have the wires in the standard EIA/TIA T568A order. (Likewise, if one end is using EIA/TIA T568B, the cable may be “straight through” if the other end is using the EIA/TIA T568B order.)

Official wiring standards would specify that this use the EIA/TIA T568A tend on both sides. However, in practice, the reality is that a cable will technically function just as well as well as long as both sides are the same. (For example, if a EIA/TIA T568B end was used on both sides, the effect would be the same.) The most likely reason this would matter is simply if one end of the cable was snipped off and re-crimped. Then it would need to match the other end, which may be easier to do when the other end is using a familiar standard (that many technicians have even memorized).

patch panel

A “patch panel” is, in essense, a bunch of Ethernet ports on a panel. Many places (particularly businesses with perhaps a dozen or more computers) will have a number of network connections (from various places, often in multiple rooms) be attached to the back side of network ports that are on a rather flat surface called a “patch panel”. A technician may then use “patch cables”, which are simply short Ethernet cables, to connect the ports to other equipment (most commonly a switch). The patch panel itself is typically mounted onto a standard “rack mount” rack.

This sort of setup allows technicians to change the wiring (such as plugging a patch panel into a different switch port), while minimizing the amount of movement/handling required for the wiring that goes into the wall. Hopefully this will reduce the likelihood of wiring within the wall breaking. Such wiring may be a challenge and expense to replace.

Multi-site Connectivity Methods

Some people get a desire to connect between two locations. Here is an overview of some such technologies/methods:

Layered Networking

Using software to create a tunnel can result in being ablet o use a “virtual private network” (“VPN”) that provides features such as privacy provided by encryption. Doing that, and relying on more public communication methods like phone lines, can be cheaper than using private lines.

Underground copper

Some things to keep in mind:

  • Determine safety before digging underground. You don't want to come into contact with underground main electrical power cables, other cabling (which may expensively disrupt other communications, and might be an electrical hazard due to having some other power cables), or pipes (which may have water flowing through them, which might start to leak upwards and create a real mess).
  • Dig deep enough. Common advice may range from 6 inches to 24 inches.
  • Don't just lay cable. Using polyvinyl chloride (“PVC”) pipes as a conduit, and then placing cables through that, can allow for repairs/adjustments without needing to re-dig the trench. This may also protect from damage by underground rocks and exposure to insects, rodents, and condensation. It may be a good idea to have conduit large enough for the current cable, more cable (e.g., a replacement), and unforeseen needs (even more cable for different technology).
  • Consider length limits. A length of 100 meters, a bit over 328 feet, is often sited as a maximum for Cat5e (or Cat5 or Cat6a) which will have Ethernet going over it. If the length of the run would be too long, find an alternative before going through the effort of digging.
  • Note where the cable is. Documentation is a great idea here.
  • Dig straight, as much as possible. Bends might be a bit harder on wires. If you bend the path of a pipe in one spot and run cable through it, every bit of the cable between that spot and one of the ends might also need to make the same bend (temporarily). Some types of cable may be less bendable, and in the future there might be a reason to consider changing cable types to a less flexible type of cable.
Fiber Optic

Fast. Not susceptible to interference from radio frequency (RF) or electro-magnetic frequency (EMF) interference. Can be reliable for kilometers/miles, instead of hundreds of meters/feet (or shorter) like limitations with common copper connections. Mostly great, with some notable limitations. One is expense. Also, repair can be more challenging, and fewer technicians have experience with that.

Copper

A common quoted limit is around 100 meters (around 300 feet), although that appears to be on the long side with some connections of that length reported to be less reliable, possibly only after a someone short period of time. If something longer is needed, then there either needs to be a powered device that re-amplifies the signal, or a different technology should be considered.

While some people may be more prone to using lower quality cable (e.g. Cat 5e) for projects within a building, these underground connections are often longer and more difficult to replace. So, using some higher quality cable may often make more sense. For instance, a quality Cat6a or Cat7 cable may be a particularly better idea than Cat5e. Using connectors that were professionally attached at a factory may be recommended over using hand-terminated cabling. (Many technicians heavily vouch for the quality of the connections they make. Experience has shown that even in such cases where the technician has very high confidence, such cables may be more prone to fail over time.)

Wi-Fi

Typicaly easier to set up than digging a trench. But some performance disadvantages compared to wired connectivity.

Other wireless

Some equipment by companies like Cambium or Ubiquiti may support longer ranges than Wi-Fi. They may be pricier. They may use some of the same frequency ranges (e.g. 2.4 GHz, 5.8 GHz) or other ranges (around 3 GHz or 60 GHz). Some of those ranges may require special liecensing. Such equipment may require line-of-sight unobstructed clear pathways of around a radius of 2 meters/6 feet (4 meters/12 feet diameter), and have a distance limitation like seven miles or so. Note that the equipment being described has been known to be used by WISPs (Wireless Internet Service Providers). (It is possible that such equipment may be priced in a way favorable to such commercial budgets.) Such equipment can be impacted by environmental effects such as ice, snow, rain, and radio frequency interference.

Longer range wireless may often involve using deployed satelite dishes. These may have different pricing requirements, and typically have more lag.