Results 1 to 12 of 12

Thread: Tcp optimezer for 4g/lte connection

  1. #1
    Junior Member
    Join Date
    May 2020
    Posts
    7

    Tcp optimezer for 4g/lte connection

    Hi everyone, I'm new here on the forum, I am inquiring about the tcp and all its values.
    I have a 4G / LTE connection at home instead of ADSL / FIBER, my router is a huawei b535-232 plus external logarithmic antennas.
    Is the TCPOptimezer program set to give the best of the right cable connections? I tried to use the recommended optimal values but instead of improving my connection, they made it worse.
    So I wanted to know the right parameters to modify to optimize my connection, I lighten screenshots.
    Sorry for my english but I live in Italy, thanks for what you do.


  2. #2
    Administrator Philip's Avatar
    Join Date
    May 1999
    Location
    Jacksonville, Florida, United States
    Posts
    10,673
    Blog Entries
    6
    The TCP Optimizer best settings depend on the download/upload speed of your connection. You can always revert to the Windows defaults as well.

    There are no specific different settings that would be required for LTE. I would suggest starting with the default settings, making small changes, then rebooting and testing. You can read what each setting does in the documentation here: https://www.speedguide.net/articles/...8-10-2012-5821

    I hope this helps, let us know what works well for your connection, and what speeds you are able to get.
    Linux is user friendly, it's just picky about its friends...
    Disclaimer: Please use caution when opening messages, my grasp on reality may have shaken loose during transmission (going on rusty memory circuits).
    ๑۩۞۩๑

  3. #3
    Junior Member
    Join Date
    May 2020
    Posts
    7
    The connection speed is good, but very fluctuating, I had attached a screenshot can't you see?
    If I set the optimal parameters in TCPOptimezer I am able to use the Nagle algorithm, why?

  4. #4
    Junior Member
    Join Date
    May 2020
    Posts
    7
    If I set the optimal values, I set ECN Capability in Default and not in Disabled, why?
    And also the Nagle algorithm all default, why?

  5. #5
    Administrator Philip's Avatar
    Join Date
    May 1999
    Location
    Jacksonville, Florida, United States
    Posts
    10,673
    Blog Entries
    6
    I did not see a screenshot.

    Some values show as default when you start the program depending on your Operating System, in case the program cannot determine their current value. This may be due to the fact that particular Windows version's default/optimal values are the same, or that the setting is network-adapter specific (you may have multiple NICs, so the program does not know which NIC's setting to show), or if the OS does not offer a reliable way to get that information about that particular setting.

    ECN is dependent on the routers your packets go through, some support it and it may be helpful, but older hardware may drop packets with the ECN bit set and introduce some packet loss. If you are not experiencing packet loss with ECN enabled you can leave it on.

    Disabling Nagle's algorithm is only recommended for gaming where you want every packet sent out without delay, not necessarily a good thing for regular web surfing and internet experience.
    Linux is user friendly, it's just picky about its friends...
    Disclaimer: Please use caution when opening messages, my grasp on reality may have shaken loose during transmission (going on rusty memory circuits).
    ๑۩۞۩๑

  6. #6

  7. #7
    Administrator Philip's Avatar
    Join Date
    May 1999
    Location
    Jacksonville, Florida, United States
    Posts
    10,673
    Blog Entries
    6
    Latency is harder to fix, as it is not caused by your own equipment, rather by other nodes on the path to a distant server. There are some settings that affect it (as TCP/IP headers change how packets are handled by routers somewhat).

    First thing would be to do a traceroute to a distant server, to see if there are any latency spikes in nodes close to you, from Windows command prompt just do something like: tracert speedguide.net
    This will show any problems/congested nodes.
    Linux is user friendly, it's just picky about its friends...
    Disclaimer: Please use caution when opening messages, my grasp on reality may have shaken loose during transmission (going on rusty memory circuits).
    ๑۩۞۩๑

  8. #8
    Junior Member
    Join Date
    May 2020
    Posts
    7
    Quote Originally Posted by Philip View Post
    Latency is harder to fix, as it is not caused by your own equipment, rather by other nodes on the path to a distant server. There are some settings that affect it (as TCP/IP headers change how packets are handled by routers somewhat).

    First thing would be to do a traceroute to a distant server, to see if there are any latency spikes in nodes close to you, from Windows command prompt just do something like: tracert speedguide.net
    This will show any problems/congested nodes.
    Thanks for the reply
    Here is the tracert to speedguide.net


    1 2 ms 1 ms 1 ms 192.168.8.1
    2 111 ms 106 ms 100 ms 172.16.128.129
    3 108 ms 101 ms 81 ms 172.16.128.4
    4 95 ms 46 ms 53 ms 172.16.13.48
    5 73 ms 101 ms 99 ms 151.7.203.129
    6 114 ms 102 ms 100 ms 151.7.203.117
    7 112 ms 100 ms 46 ms 151.7.84.225
    8 152 ms 100 ms 101 ms 151.6.1.2
    9 113 ms 97 ms 100 ms 151.6.3.21
    10 140 ms 101 ms 100 ms ipv4.decix-frankfurt.core1.fra1.he.net [80.81.192.172]
    11 116 ms 101 ms 100 ms 100ge1-1.core1.par2.he.net [72.52.92.13]
    12 257 ms 203 ms 202 ms 100ge10-2.core1.ash1.he.net [184.105.213.173]
    13 221 ms 202 ms 202 ms 100ge13-1.core1.atl1.he.net [184.105.80.162]
    14 261 ms 203 ms 169 ms e0-36.core2.jax1.he.net [184.105.213.186]
    15 187 ms 204 ms 189 ms gorack421-lc.10gigabitethernet3-5.core1.jax1.he.net [216.66.64.146]
    16 208 ms 197 ms 202 ms xe-0-0-0-1105.dist-b.jcvlfljb.jax.as19844.net [216.238.150.247]
    17 205 ms 173 ms 193 ms xe-0-1-2-1131.scolo-c10.jcvlfljb.jax.as19844.net [216.238.150.131]
    18 247 ms 204 ms 189 ms speedguide.net [68.67.73.20]

  9. #9
    Administrator Philip's Avatar
    Join Date
    May 1999
    Location
    Jacksonville, Florida, United States
    Posts
    10,673
    Blog Entries
    6
    Seems that the spike starts right from the first hop, between your 192.168.8.1 device and 172.16.128.129. This is too high, is that device your Huawei lte modem/router or what? Was it overloaded/streaming video or something else that would cause it to be congested ?
    Some causes could be internet transfer speeds close to its limit, or performing firewall/WiFi operations like automatic channel selection/scanning that would cause its CPU/RAM to be utilized close to 100%. Is it close to possible source of electromagnetic interference (EMI), like fluorescent lights, power supplies, TVs?

    Assuming that is your LTE router's IP (as it is in an internal non-routable IP range), I would try to reduce congestion/load on it to see if that helps with latency. If that is not your LTE router, but rather your ISP's node, it seems to be overloaded, I would try to complain to them about high latency to it.

    If it is not your device, you can try to change the DSCP ToS of your packets to give them higher precedence, but some ISPs don't support customer-set priorities. This is explained in more detail in this article:
    https://www.speedguide.net/articles/...-dscp-wmm-3477

  10. #10
    Junior Member
    Join Date
    May 2020
    Posts
    7
    Thanks for the reply, I read the article but it is very technical and I have not understood much ...
    Could you tell me directly how to do it? what values to set and how? Thanks again.

  11. #11
    Junior Member
    Join Date
    May 2020
    Posts
    7
    TCP options string = 020405480103030801010402
    MTU = 1392
    MTU is not fully optimized for broadband. Consider increasing your MTU to 1500 for better throughput. If you are using a router, it could be limiting your MTU regardless of Registry settings.
    MSS = 1352
    MSS is not optimized for broadband. Consider increasing your MTU value.
    Default TCP Receive Window (RWIN) = 131072
    RWIN Scaling (RFC1323) = 8 bits (scale factor: 2^8=256)
    Unscaled TCP Receive Window = 512

    In Windows 10, unless "TCP/IP Auto-Tuning" is disabled, only the Current TCP Window is displayed. Use the latest TCP Optimizer for tweaking.
    You seem to be using Google Chrome. Note that Chrome can modify the TCP Window for sockets it creates under some OSes, and therefore servers may not get your OS-assigned RWIN value. FAQ
    RWIN is not multiple of MSS. If your OS supports setting RWIN directly, consider changing it to a multiple of MSS for optimum performance.
    Other RWIN values that might work well with your current MTU/MSS:
    64896 (up to 2 Mbit lines, depending on latency. MSS * 48)
    129792 (1-5 Mbit lines, depending on latency. MSS * 48 * 2)
    259584 (2-15 Mbit lines, depending on latency. MSS * 48 * 2^2)
    519168 (10-30 Mbit lines, depending on latency. MSS * 48 * 2^3)
    1038336 (30-100 Mbit lines depending on latency. MSS * 48 * 2^4)
    bandwidth * delay product (Note this is not a speed test):

    Your current TCP Window limits you to: 5243 kbps (655 KBytes/s) @ 200ms latency
    Your current TCP Window limits you to: 2097 kbps (262 KBytes/s) @ 500ms latency
    MTU Discovery (RFC1191) = ON
    Time to live left = 112 hops
    TTL value is ok.
    Timestamps (RFC1323) = OFF
    Selective Acknowledgements (RFC2018) = ON
    IP type of service field (RFC1349) = 00000000 (0)



    SpeedGuide.net TCP Analyzer Results
    Tested on: 2020.05.21 15:38
    IP address: 151.43.xxx.x
    Client OS/browser: Windows 10 (Chrome 81.0.4044.138)

    TCP options string: 020405480103030801010402
    MSS: 1352
    MTU: 1392
    TCP Window: 131072 (not multiple of MSS)
    RWIN Scaling: 8 bits (2^8=256)
    Unscaled RWIN : 512
    Recommended RWINs: 64896, 129792, 259584, 519168, 1038336
    BDP limit (200ms): 5243kbps (655KBytes/s)
    BDP limit (500ms): 2097kbps (262KBytes/s)
    MTU Discovery: ON
    TTL: 112
    Timestamps: OFF
    SACKs: ON
    IP ToS: 00000000 (0)

  12. #12
    Administrator Philip's Avatar
    Join Date
    May 1999
    Location
    Jacksonville, Florida, United States
    Posts
    10,673
    Blog Entries
    6
    That article is a last resort technical thing... I would first try to determine what device is that second hop with 100+ms latency 172.16.128.129
    Linux is user friendly, it's just picky about its friends...
    Disclaimer: Please use caution when opening messages, my grasp on reality may have shaken loose during transmission (going on rusty memory circuits).
    ๑۩۞۩๑

Similar Threads

  1. Horrendous internet connection and connection statistics
    By Ramoge in forum Broadband Tweaks Help
    Replies: 1
    Last Post: 02-04-16, 07:12 AM
  2. Replies: 0
    Last Post: 05-02-12, 01:02 PM
  3. Tcp Optimezer
    By MetaMeta in forum Broadband Tweaks Help
    Replies: 1
    Last Post: 09-04-08, 03:46 PM
  4. Wireless connection to use a remote desktop connection
    By Niall 84 in forum alt.internet.wireless
    Replies: 0
    Last Post: 03-26-08, 06:30 PM
  5. Local Are Connection Limited Or No Connection Renew IP dont work
    By AxeOfRedemption in forum General Broadband Forum
    Replies: 0
    Last Post: 08-15-06, 11:23 PM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •