Your ISP Is Throttling or Blocking Telegram. Here Is the MTProto Fix.
Your ISP can see your Telegram traffic. It classifies it using Deep Packet Inspection and either throttles it — slowing your signals — or in some networks, restricts access entirely. Neither is acceptable when you are trading from signal rooms.
MTProto with FakeTLS removes the classification. Your ISP sees HTTPS. It cannot throttle or block what it cannot identify. Setup takes 60 seconds.
Stop the Throttle · Standard & Pro plans
What ISP throttling is and why MTProto fixes it
Your ISP uses Deep Packet Inspection to classify traffic by protocol. Telegram's MTProto protocol has a recognisable fingerprint. Once classified as Telegram, those packets are placed in a lower-priority queue — not blocked, just delayed. During peak congestion periods, that delay is 3–10 seconds.
The delay peaks during exactly the events when you need signal delivery to be fastest: NFP, FOMC, CPI. Network congestion spikes at those moments. Telegram-classified traffic waits in the lower queue. Your notification arrives after the move.
FakeTLS is a masking layer that makes an MTProto connection look exactly like a TLS 1.3 HTTPS connection to the DPI classifier. The classifier sees HTTPS, not Telegram. The traffic gets HTTPS priority. The throttle is gone.
MTProto + FakeTLS is included on Standard ($15/mo) and Pro ($25/mo). Setup: portal ? copy tg:// link ? open in Telegram ? Enable. See plans.
How to detect throttling.
Most traders who are being throttled do not know they are being throttled. The delay is intermittent — worse during news, better during quiet periods. It looks like bad luck, a slow signal room, or a platform problem. It is none of those. Here is how to confirm it.
-
1Get the Telegram notification timestamp
Note the exact time a notification appears on your screen. Use your phone's clock or a second device as reference. Do this during a news event or market open — not during a quiet midday period.
-
2Check the message timestamp in Telegram
Tap the notification and look at the message timestamp inside Telegram. Telegram logs when it received the message from the server. This is the time the message was processed by Telegram's DC — not when your notification arrived.
-
3Compare the gap
The gap between the message timestamp and your notification time is the delivery delay. Under 1–2 seconds is normal. If you consistently see 3–10 second gaps during high-volatility periods, you are being throttled.
-
4Repeat on mobile data vs Wi-Fi
Switch to mobile data and repeat the test. If the gap is dramatically smaller on mobile data, your home or office ISP is throttling on that connection specifically. If both are slow, both carriers are throttling — the fix applies to both.
-
5Repeat with MTProto proxy enabled
Set up the MTProto proxy (60 seconds from portal), then repeat the timestamp test. If the gap drops to under 1 second, the proxy is bypassing the throttle successfully. That confirms ISP DPI was the cause.
What ISP traffic shaping does to Telegram.
Every ISP manages a finite amount of network capacity. When demand exceeds supply — during busy periods — the ISP has to decide whose packets get priority. The mechanism is Quality of Service, and it runs on every major ISP's core network infrastructure.
QoS starts with traffic classification. Deep Packet Inspection hardware sits at the ISP's network nodes. It examines the shape of each connection — not the content, but the fingerprint: handshake patterns, packet timing, protocol characteristics, port combinations. Different apps produce different fingerprints. The DPI system learns to classify them.
Telegram's MTProto protocol has a consistent fingerprint. The DPI classifier sees it and marks those packets with a traffic class. That class determines their position in the queue when the network is congested.
The class ISPs assign to messaging apps — including Telegram — is typically lower priority than web browsing, but higher priority than background file transfers. In normal conditions, the queue clears fast enough that you don't notice the classification. During congestion, packets in the lower-priority class wait. Packets in the higher-priority class go first.
Your signal room message arrives at the Telegram data centre on time. It sits at the DC waiting for your device to acknowledge receipt. The acknowledgment is delayed because your device's Telegram connection is in the lower-priority queue. The message is technically received — but the push notification that wakes your screen is waiting in that queue.
"The signal was right. The ISP moved it to a slower queue. You entered late. The setup worked. Your network didn't."
Why more bandwidth doesn't help.
The instinct when something is slow is to get a faster connection. Buy the 1Gbps plan. Switch to fibre. Get a second ISP. The instinct is wrong for this problem, and understanding why matters for choosing the right fix.
The throttle is not about bandwidth. It is about queue position. Your Telegram packets are assigned to a lower-priority queue at the ISP's network node. When the queue is congested, packets are processed in priority order. Your packets wait.
A faster internet connection gets your packets to the ISP's node faster. But once they are at the node, they enter the same queue with the same priority class. A packet that arrives at the low-priority queue in 2ms rather than 5ms still waits behind all the higher-priority packets. The 3ms faster arrival changes nothing.
Bandwidth affects throughput — how much data can flow per second. Queue priority affects latency — how long each packet waits before it flows. These are different properties of a network. Trading signal delivery is a latency problem, not a throughput problem. Your signal messages are tiny — kilobytes. You have plenty of bandwidth. What you need is a different queue position.
The only way to change queue position is to change traffic classification. The only way to change traffic classification is to change the fingerprint. That is what FakeTLS does — it changes the fingerprint from Telegram to HTTPS. The DPI classifier assigns HTTPS priority. Your packets enter the higher-priority queue. The wait disappears.
The news event window: when congestion peaks at exactly the wrong time.
ISP congestion follows predictable patterns. Weekday mornings spike as people start work. Evening peak hours run from approximately 18:00 to 22:00 local time as residential users stream video and browse after work. These are the periods ISP engineers plan for.
For traders, there is an additional pattern that ISP engineers do not plan for: news event spikes. When NFP releases at 08:30 EST, every trader with Telegram open in the US sends and receives a burst of activity at the same moment. The same happens for FOMC statements, CPI releases, and major central bank decisions. The Telegram traffic profile spikes sharply in a concentrated window of seconds.
This spike compounds with general ISP congestion. The 08:30 EST release happens to coincide with peak morning congestion in the eastern US. European market open overlaps with morning ISP congestion in Europe. Asia-Pacific sessions have their own overlaps.
The result: the throttle is worst at exactly the moments it costs the most. A signal room notifying 500 traders simultaneously during NFP creates a burst of classified Telegram traffic that the DPI system puts in the lower-priority queue during a period of maximum congestion.
The first 2–3 seconds after a news release are often the entire tradeable window for the initial move. If your notification arrives 4 seconds after the signal was sent, you have missed the window. This is not bad luck. It is a predictable, preventable infrastructure failure.
Telegram congestion around a news event (schematic)
Throttle peaks in the seconds immediately after release — the only window that matters for the entry.
QoS priority queues: the mechanism in plain language.
Quality of Service is the ISP's traffic management system. At a basic level, it works like the lanes at an airport — different classes of traffic move through different lanes, and the lanes are not equal. When there is a queue, the premium lane moves first.
The lanes are defined by traffic class, which is determined by DPI classification. The ISP defines the priority rules — which traffic classes get which queue positions. The typical hierarchy for a residential ISP looks something like this:
Typical ISP QoS priority ordering
Queue positions are illustrative — specific priorities vary by ISP and region.
When the network is not congested, all queues clear quickly and the priority difference is invisible. When the network is congested — during news events, during evening peaks, during the exact moments that matter for trading — the lower-priority queue builds up and packets wait.
FakeTLS moves your Telegram traffic from the messaging app class (P3 in the diagram above) to the HTTPS class (P2). That class difference is what removes the delay. The physical network path does not change. The packet content does not change. The fingerprint changes, and the classification changes with it.
How MTProto + FakeTLS escapes the classifier.
MTProto is Telegram's native protocol — persistent TCP, AES-256 encryption, RSA-2048 key exchange with Diffie-Hellman for forward secrecy. It is a well-designed protocol. The problem is not security — it is recognisability. DPI classifiers have trained on MTProto's fingerprint.
FakeTLS is a masking extension to the MTProto proxy protocol. When FakeTLS is active, the connection opening — the handshake — is structured to exactly mimic a TLS 1.3 session. The handshake byte sequence, the record types, the timing of the initial exchange — all match what a browser would produce connecting to an HTTPS website.
The DPI classifier examines the handshake. It sees TLS 1.3. It classifies the connection as HTTPS. It moves on. It does not inspect further — classified connections are not re-inspected at depth on subsequent packets. The MTProto session runs inside what the classifier believes is a TLS tunnel.
On the proxy server side, the FakeTLS wrapper is understood and stripped. The underlying MTProto connection is established with Telegram's data centres normally. From Telegram's perspective, nothing unusual has happened — a proxy connected and is forwarding an encrypted MTProto session.
The secret in the proxy configuration — the string starting with ee — is what tells both the client and the server that FakeTLS mode is active. The ee prefix is part of Telegram's MTProto proxy specification for obfuscated mode with FakeTLS. Any proxy that claims to support FakeTLS but does not require a secret starting with ee is not correctly implementing the protocol.
The three-step fix.
From noticing the delay to having it fixed takes 60 seconds.
Go to portal.tradersproxy.com. Requires Standard ($15/mo) or Pro ($25/mo). Your MTProto credentials are in the Telegram section of the dashboard.
The portal shows a tg://proxy link with your server, port, and FakeTLS secret pre-filled. Copy the entire link. The secret starts with ee — that is correct and required for FakeTLS mode.
Paste or open the link on any device with Telegram installed. Telegram opens and shows an Enable Proxy prompt. Tap Enable. Telegram connects immediately. The throttle is bypassed from that moment forward.
For detailed platform-specific steps: full setup guide for all devices ?
What to monitor after.
Once the proxy is active, run the timestamp test again during the next news event. The gap between the message timestamp in Telegram and the notification on your screen should drop to under one second consistently — including during NFP, FOMC, and CPI windows.
The ping in Proxy settings. In Telegram ? Settings ? Privacy and Security ? Proxy, you will see a ping time for your proxy. This is the round-trip to the proxy server. A stable ping means the connection is healthy. A ping that spikes during news events would indicate proxy-side load — contact support if you see this with your portal details.
The connection indicator. The proxy should show as Connected (green dot) at all times when Telegram is running. If it drops to disconnected during backgrounding and doesn't reconnect quickly, check your device's battery optimization settings. Some phones aggressively kill background network connections — add Telegram to the battery optimization exemption list.
The next news event. The real test is the next NFP or FOMC. You should notice that notifications arrive during the event itself, not after the move. You are no longer reacting to where the price was when the signal was sent — you are reacting to the current price when the signal arrives. If the setup worked, that is the difference.
If it's still slow. Verify the proxy is connected and the secret starts with ee. If both are confirmed and you are still seeing delays, check whether your mobile carrier has a separate throttling layer. Try the same test on Wi-Fi vs mobile data with the proxy active on both. If one path is still slow, that carrier may apply a different DPI classification that requires additional troubleshooting — contact support via the portal.
Eight things worth knowing.
In most countries, yes. Net neutrality laws vary significantly. In many regions, ISPs are permitted to apply traffic shaping and QoS policies to manage network congestion, provided they don't block services outright. Throttling classified Telegram traffic during peak periods is a common, legal practice under most regulatory frameworks.
The timestamp test distinguishes them. Compare the message timestamp (when Telegram logged it as received) to the notification timestamp (when it appeared on your screen). If messages are received by Telegram near instantly but notifications arrive seconds later, the problem is on the notification delivery path — that is ISP throttling of the Telegram connection itself. If you do this test with and without the MTProto proxy and the gap disappears with the proxy active, ISP throttling was the cause.
Network congestion peaks during times when many people are online — morning commutes, lunch hours, and market open/close times often overlap. News events like NFP, FOMC, and CPI create a spike of Telegram traffic specifically among traders, which competes with general congestion for the same ISP capacity. The throttle affects all Telegram traffic, but the impact is most visible during market volatility.
Partially. A VPN can bypass DPI classification by encrypting traffic before it reaches the ISP's classifier. However, VPNs add latency overhead from encryption, route all traffic through a single tunnel (adding competition between Telegram and your terminal), and do not change the Telegram protocol fingerprint inside the tunnel on some DPI setups. MTProto + FakeTLS is the purpose-built solution for this specific problem.
Your ISP's QoS policies may throttle various classified traffic types — video streaming, P2P, VoIP, and messaging apps can all be in different priority queues. The Telegram-specific throttle only affects Telegram-classified packets. Your browser, MT5 terminal, and other apps are in separate classification categories and may not experience the same throttle.
Yes, mobile carriers apply DPI and QoS policies on cellular data networks in the same way fixed-line ISPs do. In some regions, carriers throttle messaging app traffic more aggressively during peak hours. MTProto + FakeTLS works on cellular connections in the same way it works on fixed broadband — the FakeTLS mask is protocol-level and does not depend on the access technology.
In Telegram, go to Settings ? Privacy and Security ? Proxy. Your proxy entry should show as Connected with a green dot. The secret you entered starts with 'ee' — that prefix specifically designates FakeTLS mode in Telegram's MTProto proxy protocol. If the secret starts with 'ee' and the proxy is connected, FakeTLS is active.
MTProto with FakeTLS is included on Standard ($15/mo) and Pro ($25/mo). The Starter plan does not include MTProto. To set it up, sign in to the portal, copy the tg:// link, and open it in Telegram. 60 seconds from start to connected.