diff --git a/ipacm/inc/IPACM_Wan.h b/ipacm/inc/IPACM_Wan.h
index 9acf175..9f1001a 100644
--- a/ipacm/inc/IPACM_Wan.h
+++ b/ipacm/inc/IPACM_Wan.h
@@ -370,6 +370,12 @@
 	/* handle for icmpv6 exception rule */
 	uint32_t icmpv6_exception_hdl;
 
+	/* handle for TCP FIN rule */
+	uint32_t tcp_fin_hdl;
+
+	/* handle for TCP RST rule */
+	uint32_t tcp_rst_hdl;
+
 	inline ipa_wan_client* get_client_memptr(ipa_wan_client *param, int cnt)
 	{
 	    char *ret = ((char *)param) + (wan_client_len * cnt);
@@ -636,6 +642,10 @@
 	int add_icmpv6_exception_rule();
 
 	int delete_icmpv6_exception_rule();
+
+	int add_tcp_fin_rst_exception_rule();
+
+	int delete_tcp_fin_rst_exception_rule();
 };
 
 #endif /* IPACM_WAN_H */
diff --git a/ipacm/src/IPACM_Wan.cpp b/ipacm/src/IPACM_Wan.cpp
index 140be24..464a537 100644
--- a/ipacm/src/IPACM_Wan.cpp
+++ b/ipacm/src/IPACM_Wan.cpp
@@ -1870,6 +1870,13 @@
 				IPACMERR("Failed to send ICMPv6 ex rule to modem.\n");
 				return IPACM_FAILURE;
 			}
+
+			/* send ipv4 TCP FIN filtering rule */
+			if(iptype==IPA_IP_v4 && add_tcp_fin_rst_exception_rule())
+			{
+				IPACMERR("Failed to send TCP FIN RST rule to modem.\n");
+				return IPACM_FAILURE;
+			}
 		}
 		else
 		{
@@ -4915,6 +4922,12 @@
 				IPACMERR("Failed to delete icmpv6 rule \n");
 				return IPACM_FAILURE;
 			}
+			/* Delete tcp_fin_rst rule */
+			if(delete_tcp_fin_rst_exception_rule())
+			{
+				IPACMERR("Failed to delete tcp_fin_rst rule \n");
+				return IPACM_FAILURE;
+			}
 		}
 		else
 		{
@@ -7629,7 +7642,7 @@
 	flt_rule_entry.rule.retain_hdr = 1;
 	flt_rule_entry.rule.to_uc = 0;
 	flt_rule_entry.rule.eq_attrib_type = 0;
-	flt_rule_entry.at_rear = true;
+	flt_rule_entry.at_rear = false;
 	flt_rule_entry.flt_rule_hdl = -1;
 	flt_rule_entry.status = -1;
 	flt_rule_entry.rule.action = IPA_PASS_TO_EXCEPTION;
@@ -7733,3 +7746,170 @@
 	}
 	return res;
 }
+
+int IPACM_Wan::add_tcp_fin_rst_exception_rule()
+{
+	int fd;
+	int len, res = IPACM_SUCCESS;
+	uint8_t mux_id;
+	ipa_ioc_add_flt_rule *pFilteringTable = NULL;
+
+	mux_id = ext_prop->ext[0].mux_id;
+	/* contruct filter rules to pcie modem */
+	struct ipa_flt_rule_add flt_rule_entry;
+	ipa_ioc_generate_flt_eq flt_eq;
+
+	/* construct rule */
+	IPACMDBG_H("adding MHI TCP FIN RST rule\n");
+	len = sizeof(struct ipa_ioc_add_flt_rule) + (2 * sizeof(struct ipa_flt_rule_add));
+	pFilteringTable = (struct ipa_ioc_add_flt_rule*)malloc(len);
+	if (pFilteringTable == NULL)
+	{
+		IPACMERR("Error Locate ipa_flt_rule_add memory...\n");
+		return IPACM_FAILURE;
+	}
+	memset(pFilteringTable, 0, len);
+
+	pFilteringTable->commit = 1;
+	pFilteringTable->global = false;
+	pFilteringTable->ip = IPA_IP_v4;
+	pFilteringTable->num_rules = (uint8_t)2;
+
+	/* Configuring TCP FIN RST Filtering Rule */
+	memset(&flt_rule_entry, 0, sizeof(struct ipa_flt_rule_add));
+	flt_rule_entry.rule.retain_hdr = 1;
+	flt_rule_entry.rule.to_uc = 0;
+	flt_rule_entry.rule.eq_attrib_type = 0;
+	flt_rule_entry.at_rear = false;
+	flt_rule_entry.flt_rule_hdl = -1;
+	flt_rule_entry.status = -1;
+	flt_rule_entry.rule.action = IPA_PASS_TO_EXCEPTION;
+	/*
+	 * need this since fin is last packet in an ongoing TCP connection
+	 * so it will always match the previous hash and take MHIP path
+	 */
+	flt_rule_entry.rule.hashable = false;
+
+	IPACMDBG_H("rx property attrib mask:0x%x\n", rx_prop->rx[0].attrib.attrib_mask);
+	memcpy(&flt_rule_entry.rule.attrib, &rx_prop->rx[0].attrib, sizeof(flt_rule_entry.rule.attrib));
+	flt_rule_entry.rule.attrib.attrib_mask |= IPA_FLT_PROTOCOL;
+	flt_rule_entry.rule.attrib.u.v4.protocol = (uint8_t)IPACM_FIREWALL_IPPROTO_TCP;
+
+	/* generate eq */
+	memset(&flt_eq, 0, sizeof(flt_eq));
+	memcpy(&flt_eq.attrib, &flt_rule_entry.rule.attrib, sizeof(flt_eq.attrib));
+	flt_eq.ip = IPA_IP_v4;
+
+	fd = open(IPA_DEVICE_NAME, O_RDWR);
+	if (fd < 0)
+	{
+		IPACMERR("Failed opening %s.\n", IPA_DEVICE_NAME);
+		free(pFilteringTable);
+		return IPACM_FAILURE;
+	}
+
+	if(0 != ioctl(fd, IPA_IOC_GENERATE_FLT_EQ, &flt_eq)) //define and cpy attribute to this struct
+	{
+		IPACMERR("Failed to get eq_attrib\n");
+		res = IPACM_FAILURE;
+		goto fail;
+	}
+	memcpy(&flt_rule_entry.rule.eq_attrib,
+		&flt_eq.eq_attrib,
+		sizeof(flt_rule_entry.rule.eq_attrib));
+
+	/* set the bit mask to use MEQ32_IHL offset */
+	#ifdef FEATURE_IPA_V3
+		flt_rule_entry.rule.eq_attrib.rule_eq_bitmap |= (1<<7);
+	#else
+		flt_rule_entry.rule.eq_attrib.rule_eq_bitmap |= (1<<8);
+	#endif
+
+	/* add offset to compare TCP flags */
+	flt_rule_entry.rule.eq_attrib.num_ihl_offset_meq_32 = 1;
+	flt_rule_entry.rule.eq_attrib.ihl_offset_meq_32[0].offset = 12;
+
+	/* add TCP FIN RULE */
+	flt_rule_entry.rule.eq_attrib.ihl_offset_meq_32[0].value = (((uint32_t)1)<<TCP_FIN_SHIFT);
+	flt_rule_entry.rule.eq_attrib.ihl_offset_meq_32[0].mask = (((uint32_t)1)<<TCP_FIN_SHIFT);
+	memcpy(&(pFilteringTable->rules[0]), &flt_rule_entry, sizeof(struct ipa_flt_rule_add));
+
+	/* add TCP RST rule*/
+	flt_rule_entry.rule.eq_attrib.ihl_offset_meq_32[0].value = (((uint32_t)1)<<TCP_RST_SHIFT);
+	flt_rule_entry.rule.eq_attrib.ihl_offset_meq_32[0].mask = (((uint32_t)1)<<TCP_RST_SHIFT);
+	memcpy(&(pFilteringTable->rules[1]), &flt_rule_entry, sizeof(struct ipa_flt_rule_add));
+
+	/* add rules */
+	if(false == m_filtering.AddOffloadFilteringRule(pFilteringTable, mux_id, 1))
+	{
+		IPACMERR("Failed to install WAN DL filtering table.\n");
+		res = IPACM_FAILURE;
+		goto fail;
+	}
+
+	/* save handle */
+	tcp_fin_hdl = pFilteringTable->rules[0].flt_rule_hdl;
+	tcp_rst_hdl = pFilteringTable->rules[1].flt_rule_hdl;
+
+fail:
+	close(fd);
+	if(pFilteringTable != NULL)
+	{
+		free(pFilteringTable);
+	}
+	return res;
+}
+
+int IPACM_Wan::delete_tcp_fin_rst_exception_rule()
+{
+	int len, res = IPACM_SUCCESS;
+	ipa_ioc_del_flt_rule *pFilteringTable = NULL;
+
+	struct ipa_flt_rule_del flt_rule_entry;
+
+	IPACMDBG_H("deleting MHI TCP FIN RST rule \n");
+	len = sizeof(struct ipa_ioc_del_flt_rule) + (2 * sizeof(struct ipa_flt_rule_del));
+	pFilteringTable = (struct ipa_ioc_del_flt_rule*)malloc(len);
+	if (pFilteringTable == NULL)
+	{
+		IPACMERR("Error Locate ipa_ioc_del_flt_rule memory...\n");
+		return IPACM_FAILURE;
+	}
+	memset(pFilteringTable, 0, len);
+
+	pFilteringTable->commit = 1;
+	pFilteringTable->ip = IPA_IP_v4;
+	pFilteringTable->num_hdls = (uint8_t)2;
+
+	if (tcp_fin_hdl == 0 || tcp_rst_hdl == 0)
+	{
+		IPACMERR("invalid tcp_fin_rst_hdl.\n");
+		res = IPACM_FAILURE;
+		goto fail;
+	}
+
+	memset(&flt_rule_entry, 0, sizeof(struct ipa_flt_rule_del));
+	flt_rule_entry.hdl = tcp_fin_hdl;
+
+	memcpy(&(pFilteringTable->hdl[0]), &flt_rule_entry, sizeof(struct ipa_flt_rule_del));
+
+	flt_rule_entry.hdl = tcp_rst_hdl;
+
+	memcpy(&(pFilteringTable->hdl[1]), &flt_rule_entry, sizeof(struct ipa_flt_rule_del));
+
+	if(false == m_filtering.DelOffloadFilteringRule(pFilteringTable))
+	{
+		IPACMERR("Failed to delete MHI TCP FIN RST rule.\n");
+		res = IPACM_FAILURE;
+		goto fail;
+	}
+	tcp_fin_hdl = 0;
+	tcp_rst_hdl = 0;
+
+fail:
+	if(pFilteringTable != NULL)
+	{
+		free(pFilteringTable);
+	}
+	return res;
+}