PortSIP Security Feature
Being in charge of a VoIP system in the planning to deployment stage, makes VoIP security one of your main considerations. This document presents simple and clear guidelines for PortSIP PBX, that can help you understand and make PortSIP PBX deployment more resilient to network attacks.
PortSIP PBX provides various services that use different protocols on different ports. To secure the PBX, blocked the unnecessary ports on the firewall just allow the below ports can be accessed remotely.
Service | Port | Description |
Web Portal | 8887 | Web Portal over HTTPS |
Web Portal | 8888 | Web Portal over HTTP |
Rest API | 8899 | Rest API over HTTP |
Rest API | 8900 | Rest API over HTTPS |
WebRTC | 8881 | The WebRTC client over HTTPS |
WSI | 8885 | The WebSocket Interface |
File Access | 8882;8883;8884 | The ports for file download and upload over TCP |
RTP | 25000-3499; 45000-64999 | The UDP ports for the RTP packets |
SSH | 22 | SSH port over TCP |
By default, PortSIP creates the UDP transport on 5060 and WSS transport on 5065, you can simply delete the transports and create them again with different ports. Once created the transports on new ports, don't forget to create the firewall rule by the firewalld command, and create the security group rules if deployed on the cloud platform.
We strongly suggest changing the default SSH port 22 to another port for example 10210.
By default, after the PortSIP PBX is installed, the Firewalld is enabled and all firewall rules have been configured. If installed the PBX is on Debian/Ubuntu, the default firewall UFW will be disabled.
- Installed the PortSIP PBX is in AWS/Azure/GCE, let the PBX running on a private network called VPC for AWS and GCE, for Azure, it's called VNet, then the PBX is isolated to the internet.
- In order to allow users to access PBX from the internet, a static public IP is required to assign to the PBX server.
- AWS: assign an elastic IP to the PBX EC2, and create the necessary inbound rules in the security group for services port in the above section
- Azure: associate a Public IP to the PBX VM NIC then change the IP address assignment to static, and create necessary inbound rules in the security group for service port in the above section
- GCE: in the "External IP" settings, select static external IP address to assign to the VM instance, and create the necessary VPC firewall rules for the services port in the above section
- Disable the firewalld service in PBX server: systemctl disable firewalld && systemctl stop firewalld
Important: don't stop and disable the firewalld if the PBX was deployed in the on-premise.
Separate Voice Traffic and Data Traffic for some VoIP ISPs, they provide dedicated SIP trunks that support NGN ports (Next Generation Network). NGN can separate data, voice, and video networks or any combination of the three to form a converged network.
For the on-premise deployment, the best practice is to suggest set up VLAN (Virtual Local Networks) on the PBX. VLAN can improve the call quality, but also can secure PBX. The voice traffic and data traffic can be logically separated by a VLAN switch. If one VLAN is penetrated, the other will remain secure. Also, limiting the rate of traffic to IP telephony VLANs can slow down an outside attack.
Transport Layer Security (TLS) is a mechanism for securing SIP connections. It is recommended to use TLS as PortSIP PBX SIP transport to prevent data being passed between other SIP endpoints and PortSIP PBX.
For the WebRTC client, PortSIP offers WSS transport (WebSockets over SSL/TLS). WSS is encrypted, just like HTTPS, and so protects against man-in-the-middle attacks. If the transport is secured, a range of attacks against WebSockets become unfeasible.
PortSIP PBX and PortSIP Apps support SRTP and DTLS-SRTP. SRTP extends RTP to include encryption and authentication so that all SIP and WebRTC conversations are as secure as possible. The audio and video media data is transported and protected by SRTP/DTLS-SRTP with AES-256 encryption.
PortSIP PBX provides the HTTPS and HTTP access on the port 8887 and 8888. The following are the recommended practices for securing web portal transactions and preventing unwanted access.
- Create the security rule/firewall rule to disable the HTTP access on TCP port 8888
- Disable Redirect from port 80
- Disable Redirect from port 443
- Upload the trusted SSL certificates, for example, purchase an SSL certificate from DigiCert, GeoTrust
The default username and password of PortSIP PBX administrator for Web Portal Access both are "admin". There are strongly suggest change the username and password after first logged in to the Web Portal.
- Click the left menu "Profile", on the "General" page, enter a difficult to guess username then click the "Apply" button to change the default username
- Click on the "avatar" in the upper right corner, choose the "Change Password" menu, then enter the current password and new password, the new password must meet all the following requirements
- Include at least one number
- Include at least one lower-case letter
- Include at least one upper-case letter
- The password size must in 8 - 16 characters
After a tenant was created, the PBX administrator has been setting up the username and password for the tenant manager.
There are strongly suggest change the username and password after the tenant manager first logged in to the Web Portal.
- Click the left menu "Profile", on the "General" page, enter a difficult to guess username then click the "Apply" button to change the default username
- Click on the "avatar" in the upper right corner, choose the "Change Password" menu, then enter the current password and new password, the new password must meet all the following requirements
- Include at least one number
- Include at least one lower-case letter
- Include at least one upper-case letter
- The password size must in 8 - 16 characters
After a tenant was created, the tenant manager has been setting up the extension number and password for the extension, there have two passwords with an extension.
- SIP Password. It's using for the IP Phone, Softphone, WebRTC client to register to PortSIP PBX
- Web Password. It's using for the extension sign the PBX Web Portal to checking voicemail, recording, CDR
- Both SIP Password and Web Password must meet all the following requirements
- Include at least one number
- Include at least one lower-case letter
- Include at least one upper-case letter
- The password size must in 8 - 16 characters
After the PortSIP PBX administrator signed in Web Portal, there have some settings that allow to secure the login security for tenant managers, tenants, and extensions.
- Click the left menu "Advanced > Security", on the "Web Login" page, set the maximum number of login tries on the "Web Login" page, and the user IP will be blocked if the number of failed login attempts exceeds the allowed times
- Set the period of an IP block, a blocked IP will be removed after this time
PortSIP PBX provides security features with the main purpose to block any malicious attacks targeted to the PortSIP PBX in case the administrator has not taken necessary precautions at the firewall level. It works by detecting and blocking packet floods / DoS attacks or brute force dictionary attacks within the scope of identifying and cracking the extension number and the password.
Click the left menu "Advanced > Security", on the "Anti Hacking" page, there shows the main interface of the PortSIP PBX Anti Hacking configurations.
This is a time interval in seconds when counting starts but no action is enforced. To disable security, set it to a higher value.
This is the protection in case the attacker tries to use a dictionary attack to guess the password set for a particular extension.
To do this the attacker has to send numerous requests and after the server sends a “Proxy Authentication Required” message the attacker will send a request with authentication. With this feature, the attacker can only send 50 requests in an attempt to crack the password. If an IP Address spams PortSIP with 50 wrong Authentication attempts in “Detection Period”, that IP address will be blocked and put in the blacklist for the time specified in the “SIP Blacklist time interval” parameter, by default 1 hour.
D.O.S attacks can send REGISTER/INVITE requests but do not reply to Challenge (407). Configure the amount of "fake" requests that PortSIP PBX will accept per IP Address. If this value is exceeded in the "Detection Period" interval the source IP address is put in the Blacklist. IP will remain blacklisted till the "SIP Blacklist time interval" expires, by default 1 hour.
This is the 2nd layer of protection. Here you can specify how many packets can be sent from a unique source IP address. The default value is 2000 packets per second. If an IP Address is sending more than 200 packets per second, it means that there is something wrong. At this point, the attacker IP will be blocked until the "Level 2 blacklist time interval" expires.
This is the 1st layer in packets per second. If an IP sends more packets than the amount specified per second, it will get blacklisted for the “Level 1 blacklist time interval”. By default value is 500 packets per second.
At this layer, once that packet rate exceeds the rating, the blacklist is enforced, the user IP will get blacklisted for the “Level 1 blacklist time interval”.
Once an IP address was blocked due to the above L1/L2 rules, it will display in the menu "Blacklist and Codes > IP Blacklist", from which you can add it into “Whitelist” manually.
All extensions are assigned to the "Default" extension group by default, the tenant manager can change the default user group settings to secure the extension.
Click "Call Manager > Extension Groups" on the left menu, then double-click the listed group to see the settings.
- Allow Paging/Intercom: if selected, the group members have the ability to make the Paging/Intercom call
- Allow external call: if selected, the group members have the ability to make the outbound call to PSTN
- Allow Access to Web Portal: if selected, the group members have the ability to sign in the Web Portal
- Enable Call Group Members: if selected, the group members have the ability call to other group members
PortSIP PBX allows you to whitelist and blacklist IP addresses. All traffic originating from whitelisted IP addresses will be allowed unchecked by the anti-hacking features. All traffic originating from blacklisted IP addresses will be dropped immediately and silently.
Assume a remote office is connected to the PortSIP PBX. The public IP address of the remote office is 103.224.182.210. This IP address's traffic is safe to trust. Follow the settings below to add this IP address to a whitelist.
- Click on "Blacklist and Codes" > "IP Blacklist"
- Click "Add" to add an entry
- Enter the IP address that you want to allow – in this example it should be 103.224.182.210 (you can also enter the IP 103.224.182.210.0 and choose a Subnet Mask to allow an IP range)
- Choose "Allow" for the "Action" dropdown
- Add a description for the IP address, for example, "My Remote office"
- Click "Apply", the allow entry will be created in the IP Blacklist page for the whitelisted IP address. All traffic originated from this IP address will not be checked and the anti-hacking algorithms will not come into effect
Let us look at another scenario. Assume that there is a distributed attack coming from the following IP addresses – 41.202.160.2 and 41.202.191.5. These two IP addresses have already been blacklisted by PortSIP PBX’s anti-hacking auto-detection mechanisms. You would, however, want to blacklist all the range, since you are sure that you will never get any traffic from these IP addresses. In this case, we will blacklist the whole range from 41.202.0.0 to 41.202.255.255, i.e. all the IP addresses that started with 41.202. 1.
- Click on "Blacklist and Codes" > "IP Blacklist"
- Click "Add" to add an entry
- In the “IP address” enter the first address of the network range you want to block. For this example, we will enter 41.202.0.0
- Since we want to block all IP addresses started with 41.202, we will select a Subnet Mask of 255.255.0.0. The range of IP addresses contained in this mask will be displayed below
- Set Action to “Permanently block“
- Enter a Description for this entry to help you remember why you added this entry, for example, “Anti D.O.S attack coming from 41.202.x.x”
- Click “Apply“. A blocked entry will be created on the IP Blacklist page. All traffic coming from this IP address range will be checked, anti-hacking algorithms will come into effect and all packets from this IP Address range will be completely dropped and ignored
- The PortSIP Blacklist/Whitelist mechanism does not conform to a replacement of firewall. It merely provides a defense mechanism to help differentiate traffic trustable, and traffic not trustworthy. If, for example, you want to block all traffic to your network and allow only your VoIP Provider IP address, you need to set this up on your firewall
When configuring a range of IP addresses in the Blacklist, you should also ensure that the range does not include the IP address of which the PBX is installed.
SIP Trunking is often a Peer to Peer connection for the primary use of delivering PSTN connectivity over VoIP. SIP Trunking is delivered over a couple of different methods, Internet Telephony Service Providers (ITSP) deliver SIP Trunking over the Internet and Managed Service Providers deliver SIP Trunking over the dedicated carriers WAN connections. The application of security solutions involves providing a Firewall in combination with an IP-PBX that is used to define the Peer to Peer relationship at various networks and VoIP application layers, and also ensuring signaling and media are secure as well.
Register Based Authentication: Many SIP Trunk Service Providers will require a level of Authentication within the SIP Trunk. The Service Provider requires Registration Authentication and Call Initiation Authentication from the PBX. When the PBX initiates a call to the Service Provider, the PBX must provide Authentication within the SIP Protocol for the Service Provider to accept and process the call.
IP Based Authentication: Because some SIP Trunk Service Providers do not support the SIP REGISTER method, you'll need to set up Trunk as the "IP Based" and add Trunk IP addresses as trusted peers in PBX, then the PBX is to accept SIP traffic from trunk IP do not challenge for authentication credentials.
PortSIP PBX supports both Register Based and IP Based Authentication Trunks, but the IP Based Authentication trunk is strongly recommended, it's more secure.
PortSIP PBX is also supporting accept the Trunk/E1/T1 gateway registration. For example, if an E1/T1 gateway is located in a local LAN but the PBX is in the cloud, we can create a "Accept Register" Trunk in PortSIP PBX, set the username and password, and the E1/T1 gateway will be able to use that username and password register to the PortSIP PBX, the PBX only allows make & accept calls with E1/T1 gateway after successfully authorized.
When creating the outbound rule in the PortSIP PBX, you will need to consider outbound rule permission for different users.
You can create the outbound rule using the provided called number prefix, called number length, and caller belonged extension groups.
For example, you can set up outbound rules as below.
- Outbound rule for local calls
- Create an outbound rule and select the trunk that is least-cost for local calls, and set the outbound rule permission for an extension group or for multiple extension groups, the extension who in the groups will have the ability to make the local calls
- Outbound rule for long-distance calls
- Create an outbound rule and select the trunk that is least-cost for national calls, and set the outbound rule permission for an extension group or multiple extension groups, the extension who in the groups will have the ability to make the national calls
- Outbound rule for international calls
- Create an outbound rule and select the trunk that is least-cost for international calls, and set the outbound rule permission for an extension group or multiple extension groups, the extension who in the groups will have the ability to make the international calls.
- Office hours for the outbound rule
- PortSIP PBX allows specified office hours for an outbound rule, once set, the outbound rule will be unavailable and no one can make the call on it when outside of those hours.
The tenant manager can simply disable the outbound calls authorization for specific extension groups.
Click "Call Manager > Extension Groups" on the left menu, then double-click the listed group to see the settings.
- Allow Paging/Intercom: if this option is not chosen, all members of the group will be unable to make outbound calls
Last modified 2yr ago