diff --git a/net/tcp/Kconfig b/net/tcp/Kconfig
index 325f7b79446..df88824b8f1 100644
--- a/net/tcp/Kconfig
+++ b/net/tcp/Kconfig
@@ -79,6 +79,33 @@ config NET_MAX_LISTENPORTS
 	---help---
 		Maximum number of listening TCP/IP ports (all tasks).  Default: 20
 
+config NET_TCP_FAST_RETRANSMIT_WATERMARK
+	int "WaterMark to trigger Fast Retransmission"
+	default 3
+	---help---
+		RFC2001:
+		3.  Fast Retransmit
+			Modifications to the congestion avoidance algorithm were proposed in
+			1990 [3].  Before describing the change, realize that TCP may
+			generate an immediate acknowledgment (a duplicate ACK) when an out-
+			of-order segment is received (Section 4.2.2.21 of [1], with a note
+			that one reason for doing so was for the experimental fast-
+			retransmit algorithm).  This duplicate ACK should not be delayed.
+			The purpose of this duplicate ACK is to let the other end know that a
+			segment was received out of order, and to tell it what sequence
+			number is expected.
+
+			Since TCP does not know whether a duplicate ACK is caused by a lost
+			segment or just a reordering of segments, it waits for a small number
+			of duplicate ACKs to be received.  It is assumed that if there is
+			just a reordering of the segments, there will be only one or two
+			duplicate ACKs before the reordered segment is processed, which will
+			then generate a new ACK.  If three or more duplicate ACKs are
+			received in a row, it is a strong indication that a segment has been
+			lost.  TCP then performs a retransmission of what appears to be the
+			missing segment, without waiting for a retransmission timer to
+			expire.
+
 config NET_TCP_NOTIFIER
 	bool "Support TCP notifications"
 	default n
diff --git a/net/tcp/tcp.h b/net/tcp/tcp.h
index 044869fa54f..864164d7bc6 100644
--- a/net/tcp/tcp.h
+++ b/net/tcp/tcp.h
@@ -79,6 +79,7 @@
 #  define TCP_WBPKTLEN(wrb)          ((wrb)->wb_iob->io_pktlen)
 #  define TCP_WBSENT(wrb)            ((wrb)->wb_sent)
 #  define TCP_WBNRTX(wrb)            ((wrb)->wb_nrtx)
+#  define TCP_WBNACK(wrb)            ((wrb)->wb_nack)
 #  define TCP_WBIOB(wrb)             ((wrb)->wb_iob)
 #  define TCP_WBCOPYOUT(wrb,dest,n)  (iob_copyout(dest,(wrb)->wb_iob,(n),0))
 #  define TCP_WBCOPYIN(wrb,src,n) \
@@ -299,6 +300,7 @@ struct tcp_wrbuffer_s
   uint16_t   wb_sent;      /* Number of bytes sent from the I/O buffer chain */
   uint8_t    wb_nrtx;      /* The number of retransmissions for the last
                             * segment sent */
+  uint8_t    wb_nack;      /* The number of ack count */
   struct iob_s *wb_iob;    /* Head of the I/O buffer chain */
 };
 #endif
diff --git a/net/tcp/tcp_send_buffered.c b/net/tcp/tcp_send_buffered.c
index 040ad561831..3736b1beac3 100644
--- a/net/tcp/tcp_send_buffered.c
+++ b/net/tcp/tcp_send_buffered.c
@@ -326,6 +326,7 @@ static uint16_t psock_send_eventhandler(FAR struct net_driver_s *dev,
 {
   FAR struct tcp_conn_s *conn = (FAR struct tcp_conn_s *)pvconn;
   FAR struct socket *psock = (FAR struct socket *)pvpriv;
+  bool rexmit = false;
 
   /* Check for a loss of connection */
 
@@ -482,6 +483,28 @@ static uint16_t psock_send_eventhandler(FAR struct net_driver_s *dev,
                           wrb, TCP_WBSEQNO(wrb), TCP_WBPKTLEN(wrb));
                 }
             }
+          else if (ackno == TCP_WBSEQNO(wrb))
+            {
+              /* Duplicate ACK? Retransmit data if need */
+
+              TCP_WBNACK(wrb)++;
+
+              if (TCP_WBNACK(wrb) ==
+                  CONFIG_NET_TCP_FAST_RETRANSMIT_WATERMARK)
+                {
+                  /* Do fast retransmit */
+
+                  rexmit = true;
+                }
+              else if ((TCP_WBNACK(wrb) >
+                       CONFIG_NET_TCP_FAST_RETRANSMIT_WATERMARK) &&
+                       TCP_WBNACK(wrb) == sq_count(&conn->unacked_q) - 1)
+                {
+                  /* Reset the duplicate ack counter */
+
+                  TCP_WBNACK(wrb) = 0;
+                }
+            }
         }
 
       /* A special case is the head of the write_q which may be partially
@@ -521,6 +544,11 @@ static uint16_t psock_send_eventhandler(FAR struct net_driver_s *dev,
   /* Check if we are being asked to retransmit data */
 
   else if ((flags & TCP_REXMIT) != 0)
+    {
+      rexmit = true;
+    }
+
+  if (rexmit)
     {
       FAR struct tcp_wrbuffer_s *wrb;
       FAR sq_entry_t *entry;
diff --git a/net/tcp/tcp_wrbuffer.c b/net/tcp/tcp_wrbuffer.c
index 82df41dc4bf..03dff66d449 100644
--- a/net/tcp/tcp_wrbuffer.c
+++ b/net/tcp/tcp_wrbuffer.c
@@ -260,6 +260,10 @@ void tcp_wrbuffer_release(FAR struct tcp_wrbuffer_s *wrb)
       iob_free_chain(wrb->wb_iob, IOBUSER_NET_TCP_WRITEBUFFER);
     }
 
+  /* Reset the ack counter */
+
+  TCP_WBNACK(wrb) = 0;
+
   /* Then free the write buffer structure */
 
   sq_addlast(&wrb->wb_node, &g_wrbuffer.freebuffers);